v0.10.2

idk forgot to pull

.vscode/tasks.json

@@ -8,8 +8,10 @@
         },
         {
             "label": "no_dbg",
-            "command": "${config:python.pythonPath} -m copyparty -ed -emp -e2dsa -e2ts -a ed:wark -v srv::r:aed:cnodupe -v dist:dist:r ;exit 1",
-            "type": "shell"
+            "type": "shell",
+            "command": "${config:python.pythonPath} -m copyparty -ed -emp -e2dsa -e2ts -a ed:wark -v srv::r:aed:cnodupe -v dist:dist:r ;exit 1"
+            // -v ~/Music/mt:mt:r:cmtp=.bpm=~/dev/copyparty/bin/mtag/audio-bpm.py:cmtp=key=~/dev/copyparty/bin/mtag/audio-key.py:ce2tsr 
+            // -v ~/Music/mt:mt:r:cmtp=.bpm=~/dev/copyparty/bin/mtag/audio-bpm.py:ce2tsr
         }
     ]
 }

README.md

@@ -13,6 +13,31 @@ turn your phone or raspi into a portable file server with resumable uploads/down
 * code standard: `black`
 
 
+## readme toc
+
+* top
+    * [quickstart](#quickstart)
+    * [notes](#notes)
+    * [status](#status)
+* [bugs](#bugs)
+* [usage](#usage)
+    * [zip downloads](#zip-downloads)
+* [searching](#searching)
+    * [search configuration](#search-configuration)
+    * [metadata from audio files](#metadata-from-audio-files)
+    * [file parser plugins](#file-parser-plugins)
+    * [complete examples](#complete-examples)
+* [client examples](#client-examples)
+* [dependencies](#dependencies)
+    * [optional gpl stuff](#optional-gpl-stuff)
+* [sfx](#sfx)
+    * [sfx repack](#sfx-repack)
+* [install on android](#install-on-android)
+* [dev env setup](#dev-env-setup)
+* [how to release](#how-to-release)
+* [todo](#todo)
+
+
 ## quickstart
 
 download [copyparty-sfx.py](https://github.com/9001/copyparty/releases/latest/download/copyparty-sfx.py) and you're all set!
@@ -48,7 +73,7 @@ you may also want these, especially on servers:
   * ☑ symlink/discard existing files (content-matching)
 * download
   * ☑ single files in browser
-  * ✖ folders as zip files
+  * ☑ folders as zip / tar files
   * ☑ FUSE client (read-only)
 * browser
   * ☑ tree-view
@@ -71,9 +96,37 @@ summary: it works! you can use it! (but technically not even close to beta)
 
 * Windows: python 3.7 and older cannot read tags with ffprobe, so use mutagen or upgrade
 * Windows: python 2.7 cannot index non-ascii filenames with `-e2d`
+* Windows: python 2.7 cannot handle filenames with mojibake
 * probably more, pls let me know
 
 
+# usage
+
+the browser has the following hotkeys
+* `0..9` jump to 10%..90%
+* `U/O` skip 10sec back/forward
+* `J/L` prev/next song
+* `I/K` prev/next folder
+* `P` parent folder
+
+
+## zip downloads
+
+the `zip` link next to folders can produce various types of zip/tar files using these alternatives in the browser settings tab:
+
+| name | url-suffix | description |
+|--|--|--|
+| `tar` | `?tar` | plain gnutar, works great with `curl \| tar -xv` |
+| `zip` | `?zip=utf8` | works everywhere, glitchy filenames on win7 and older |
+| `zip_dos` | `?zip` | traditional cp437 (no unicode) to fix glitchy filenames |
+| `zip_crc` | `?zip=crc` | cp437 with crc32 computed early for truly ancient software |
+
+* hidden files (dotfiles) are excluded unless `-ed`
+  * the up2k.db is always excluded
+* `zip_crc` will take longer to download since the server has to read each file twice
+  * please let me know if you find a program old enough to actually need this
+
+
 # searching
 
 when started with `-e2dsa` copyparty will scan/index all your files. This avoids duplicates on upload, and also makes the volumes searchable through the web-ui:
@@ -100,19 +153,24 @@ through arguments:
 * `-e2tsr` deletes all existing tags, so a full reindex
 
 the same arguments can be set as volume flags, in addition to `d2d` and `d2t` for disabling:
-* `-v ~/music::ce2dsa:ce2tsr` does a full reindex of everything on startup
-* `-v ~/music::cd2d` disables **all** indexing, even if any `-e2*` are on
-* `-v ~/music::cd2t` disables all `-e2t*` (tags), does not affect `-e2d*`
+* `-v ~/music::r:ce2dsa:ce2tsr` does a full reindex of everything on startup
+* `-v ~/music::r:cd2d` disables **all** indexing, even if any `-e2*` are on
+* `-v ~/music::r:cd2t` disables all `-e2t*` (tags), does not affect `-e2d*`
 
 `e2tsr` is probably always overkill, since `e2ds`/`e2dsa` would pick up any file modifications and cause `e2ts` to reindex those
 
+
+## metadata from audio files
+
 `-mte` decides which tags to index and display in the browser (and also the display order), this can be changed per-volume:
-* `-v ~/music::cmte=title,artist` indexes and displays *title* followed by *artist*
+* `-v ~/music::r:cmte=title,artist` indexes and displays *title* followed by *artist*
 
 if you add/remove a tag from `mte` you will need to run with `-e2tsr` once to rebuild the database, otherwise only new files will be affected
 
 `-mtm` can be used to add or redefine a metadata mapping, say you have media files with `foo` and `bar` tags and you want them to display as `qux` in the browser (preferring `foo` if both are present), then do `-mtm qux=foo,bar` and now you can `-mte artist,title,qux`
 
+tags that start with a `.` such as `.bpm` and `.dur`(ation) indicate numeric value
+
 see the beautiful mess of a dictionary in [mtag.py](https://github.com/9001/copyparty/blob/master/copyparty/mtag.py) for the default mappings (should cover mp3,opus,flac,m4a,wav,aif,)
 
 `--no-mutagen` disables mutagen and uses ffprobe instead, which...
@@ -122,6 +180,21 @@ see the beautiful mess of a dictionary in [mtag.py](https://github.com/9001/copy
 * more importantly runs ffprobe on incoming files which is bad if your ffmpeg has a cve
 
 
+## file parser plugins
+
+copyparty can invoke external programs to collect additional metadata for files using `mtp` (as argument or volume flag), there is a default timeout of 30sec
+
+* `-mtp .bpm=~/bin/audio-bpm.py` will execute `~/bin/audio-bpm.py` with the audio file as argument 1 to provide the `.bpm` tag, if that does not exist in the audio metadata
+* `-mtp key=f,t5,~/bin/audio-key.py` uses `~/bin/audio-key.py` to get the `key` tag, replacing any existing metadata tag (`f,`), aborting if it takes longer than 5sec (`t5,`)
+* `-v ~/music::r:cmtp=.bpm=~/bin/audio-bpm.py:cmtp=key=f,t5,~/bin/audio-key.py` both as a per-volume config wow this is getting ugly
+
+
+## complete examples
+
+* read-only music server with bpm and key scanning  
+  `python copyparty-sfx.py -v /mnt/nas/music:/music:r -e2dsa -e2ts -mtp .bpm=f,audio-bpm.py -mtp key=f,audio-key.py`
+
+
 # client examples
 
 * javascript: dump some state into a file (two separate examples)
@@ -158,6 +231,13 @@ copyparty returns a truncated sha512sum of your PUT/POST as base64; you can gene
 * `Pillow` (requires py2.7 or py3.5+)
 
 
+## optional gpl stuff
+
+some bundled tools have copyleft dependencies, see [./bin/#mtag](bin/#mtag)
+
+these are standalone and will never be imported / evaluated by copyparty
+
+
 # sfx
 
 currently there are two self-contained binaries:

bin/README.md

@@ -1,4 +1,4 @@
-# copyparty-fuse.py
+# [`copyparty-fuse.py`](copyparty-fuse.py)
 * mount a copyparty server as a local filesystem (read-only)
 * **supports Windows!** -- expect `194 MiB/s` sequential read
 * **supports Linux** -- expect `117 MiB/s` sequential read
@@ -29,7 +29,7 @@ you could replace winfsp with [dokan](https://github.com/dokan-dev/dokany/releas
 
 
 
-# copyparty-fuse🅱️.py
+# [`copyparty-fuse🅱️.py`](copyparty-fuseb.py)
 * mount a copyparty server as a local filesystem (read-only)
 * does the same thing except more correct, `samba` approves
 * **supports Linux** -- expect `18 MiB/s` (wait what)
@@ -37,5 +37,11 @@ you could replace winfsp with [dokan](https://github.com/dokan-dev/dokany/releas
 
 
 
-# copyparty-fuse-streaming.py
+# [`copyparty-fuse-streaming.py`](copyparty-fuse-streaming.py)
 * pretend this doesn't exist
+
+
+
+# [`mtag/`](mtag/)
+* standalone programs which perform misc. file analysis
+* copyparty can Popen programs like these during file indexing to collect additional metadata

bin/mtag/README.md

@@ -0,0 +1,34 @@
+standalone programs which take an audio file as argument
+
+some of these rely on libraries which are not MIT-compatible
+
+* [audio-bpm.py](./audio-bpm.py) detects the BPM of music using the BeatRoot Vamp Plugin; imports GPL2
+* [audio-key.py](./audio-key.py) detects the melodic key of music using the Mixxx fork of keyfinder; imports GPL3
+
+
+# dependencies
+
+run [`install-deps.sh`](install-deps.sh) to build/install most dependencies required by these programs (supports windows/linux/macos)
+
+*alternatively* (or preferably) use packages from your distro instead, then you'll need at least these:
+
+* from distro: `numpy vamp-plugin-sdk beatroot-vamp mixxx-keyfinder ffmpeg`
+* from pypy: `keyfinder vamp`
+
+
+# usage from copyparty
+
+`copyparty -e2dsa -e2ts -mtp key=f,audio-key.py -mtp .bpm=f,audio-bpm.py`
+
+* `f,` makes the detected value replace any existing values
+* the `.` in `.bpm` indicates numeric value
+* assumes the python files are in the folder you're launching copyparty from, replace the filename with a relative/absolute path if that's not the case
+* `mtp` modules will not run if a file has existing tags in the db, so clear out the tags with `-e2tsr` the first time you launch with new `mtp` options
+
+
+## usage with volume-flags
+
+instead of affecting all volumes, you can set the options for just one volume like so:
+```
+copyparty -v /mnt/nas/music:/music:r:cmtp=key=f,audio-key.py:cmtp=.bpm=f,audio-bpm.py:ce2dsa:ce2ts
+```

bin/mtag/audio-bpm.py

@@ -0,0 +1,69 @@
+#!/usr/bin/env python
+
+import os
+import sys
+import vamp
+import tempfile
+import numpy as np
+import subprocess as sp
+
+from copyparty.util import fsenc
+
+"""
+dep: vamp
+dep: beatroot-vamp
+dep: ffmpeg
+"""
+
+
+def det(tf):
+    # fmt: off
+    sp.check_call([
+        "ffmpeg",
+        "-nostdin",
+        "-hide_banner",
+        "-v", "fatal",
+        "-ss", "13",
+        "-y", "-i", fsenc(sys.argv[1]),
+        "-ac", "1",
+        "-ar", "22050",
+        "-t", "300",
+        "-f", "f32le",
+        tf
+    ])
+    # fmt: on
+
+    with open(tf, "rb") as f:
+        d = np.fromfile(f, dtype=np.float32)
+        try:
+            # 98% accuracy on jcore
+            c = vamp.collect(d, 22050, "beatroot-vamp:beatroot")
+            cl = c["list"]
+        except:
+            # fallback; 73% accuracy
+            plug = "vamp-example-plugins:fixedtempo"
+            c = vamp.collect(d, 22050, plug, parameters={"maxdflen": 40})
+            print(c["list"][0]["label"].split(" ")[0])
+            return
+
+        # throws if detection failed:
+        bpm = float(cl[-1]["timestamp"] - cl[1]["timestamp"])
+        bpm = round(60 * ((len(cl) - 1) / bpm), 2)
+        print(f"{bpm:.2f}")
+
+
+def main():
+    with tempfile.NamedTemporaryFile(suffix=".pcm", delete=False) as f:
+        f.write(b"h")
+        tf = f.name
+
+    try:
+        det(tf)
+    except:
+        pass
+    finally:
+        os.unlink(tf)
+
+
+if __name__ == "__main__":
+    main()

bin/mtag/audio-key.py

@@ -0,0 +1,18 @@
+#!/usr/bin/env python
+
+import sys
+import keyfinder
+
+"""
+dep: github/mixxxdj/libkeyfinder
+dep: pypi/keyfinder
+dep: ffmpeg
+
+note: cannot fsenc
+"""
+
+
+try:
+    print(keyfinder.key(sys.argv[1]).camelot())
+except:
+    pass

bin/mtag/install-deps.sh

@@ -0,0 +1,265 @@
+#!/bin/bash
+set -e
+
+
+# install dependencies for audio-*.py
+#
+# linux: requires {python3,ffmpeg,fftw}-dev py3-{wheel,pip} py3-numpy{,-dev} vamp-sdk-dev patchelf
+# win64: requires msys2-mingw64 environment
+# macos: requires macports
+#
+# has the following manual dependencies, especially on mac:
+#   https://www.vamp-plugins.org/pack.html
+#
+# installs stuff to the following locations:
+#   ~/pe/
+#   whatever your python uses for --user packages
+#
+# does the following terrible things:
+#   modifies the keyfinder python lib to load the .so in ~/pe
+
+
+linux=1
+
+win=
+[ ! -z "$MSYSTEM" ] || [ -e /msys2.exe ] && {
+	[ "$MSYSTEM" = MINGW64 ] || {
+		echo windows detected, msys2-mingw64 required
+		exit 1
+	}
+	pacman -S --needed mingw-w64-x86_64-{ffmpeg,python,python-pip,vamp-plugin-sdk}
+	win=1
+	linux=
+}
+
+mac=
+[ $(uname -s) = Darwin ] && {
+	#pybin="$(printf '%s\n' /opt/local/bin/python* | (sed -E 's/(.*\/[^/0-9]+)([0-9]?[^/]*)$/\2 \1/' || cat) | (sort -nr || cat) | (sed -E 's/([^ ]*) (.*)/\2\1/' || cat) | grep -E '/(python|pypy)[0-9\.-]*$' | head -n 1)"
+	pybin=/opt/local/bin/python3.9
+	[ -e "$pybin" ] || {
+		echo mac detected, python3 from macports required
+		exit 1
+	}
+	pkgs='ffmpeg python39 py39-wheel'
+	ninst=$(port installed | awk '/^  /{print$1}' | sort | uniq | grep -E '^('"$(echo "$pkgs" | tr ' ' '|')"')$' | wc -l)
+	[ $ninst -eq 3 ] || {
+		sudo port install $pkgs
+	}
+	mac=1
+	linux=
+}
+
+hash -r
+
+[ $mac ] || {
+	command -v python3 && pybin=python3 || pybin=python
+}
+
+$pybin -m pip install --user numpy
+
+
+command -v gnutar && tar() { gnutar "$@"; }
+command -v gtar && tar() { gtar "$@"; }
+command -v gsed && sed() { gsed "$@"; }
+
+
+need() {
+	command -v $1 >/dev/null || {
+		echo need $1
+		exit 1
+	}
+}
+need cmake
+need ffmpeg
+need $pybin
+#need patchelf
+
+
+td="$(mktemp -d)"
+cln() {
+	rm -rf "$td"
+}
+trap cln EXIT
+cd "$td"
+pwd
+
+
+dl_text() {
+	command -v curl >/dev/null && exec curl "$@"
+	exec wget -O- "$@"
+}
+dl_files() {
+	local yolo= ex=
+	[ $1 = "yolo" ] && yolo=1 && ex=k && shift
+	command -v curl >/dev/null && exec curl -${ex}JOL "$@"
+	
+	[ $yolo ] && ex=--no-check-certificate
+	exec wget --trust-server-names $ex "$@"
+}
+export -f dl_files
+
+
+github_tarball() {
+	dl_text "$1" |
+	tee json |
+	(
+		# prefer jq if available
+		jq -r '.tarball_url' ||
+
+		# fallback to awk (sorry)
+		awk -F\" '/"tarball_url": "/ {print$4}'
+	) |
+	tee /dev/stderr |
+	tr -d '\r' | tr '\n' '\0' |
+	xargs -0 bash -c 'dl_files "$@"' _
+}
+
+
+gitlab_tarball() {
+	dl_text "$1" |
+	tee json |
+	(
+		# prefer jq if available
+		jq -r '.[0].assets.sources[]|select(.format|test("tar.gz")).url' ||
+
+		# fallback to abomination
+		tr \" '\n' | grep -E '\.tar\.gz$' | head -n 1
+	) |
+	tee /dev/stderr |
+	tr -d '\r' | tr '\n' '\0' |
+	tee links |
+	xargs -0 bash -c 'dl_files "$@"' _
+}
+
+
+install_keyfinder() {
+	# windows support:
+	#   use msys2 in mingw-w64 mode
+	#   pacman -S --needed mingw-w64-x86_64-{ffmpeg,python}
+	
+	github_tarball https://api.github.com/repos/mixxxdj/libkeyfinder/releases/latest
+
+	tar -xf mixxxdj-libkeyfinder-*
+	rm -- *.tar.gz
+	cd mixxxdj-libkeyfinder*
+	
+	h="$HOME"
+	so="lib/libkeyfinder.so"
+	memes=()
+
+	[ $win ] &&
+		so="bin/libkeyfinder.dll" &&
+		h="$(printf '%s\n' "$USERPROFILE" | tr '\\' '/')" &&
+		memes+=(-G "MinGW Makefiles" -DBUILD_TESTING=OFF)
+	
+	[ $mac ] &&
+		so="lib/libkeyfinder.dylib"
+
+	cmake -DCMAKE_INSTALL_PREFIX="$h/pe/keyfinder" "${memes[@]}" -S . -B build
+	cmake --build build --parallel $(nproc || echo 4)
+	cmake --install build
+
+	libpath="$h/pe/keyfinder/$so"
+	[ $linux ] && [ ! -e "$libpath" ] &&
+		so=lib64/libkeyfinder.so
+	
+	libpath="$h/pe/keyfinder/$so"
+	[ -e "$libpath" ] || {
+		echo "so not found at $sop"
+		exit 1
+	}
+	
+	# rm -rf /Users/ed/Library/Python/3.9/lib/python/site-packages/*keyfinder*
+	CFLAGS="-I$h/pe/keyfinder/include -I/opt/local/include" \
+	LDFLAGS="-L$h/pe/keyfinder/lib -L$h/pe/keyfinder/lib64 -L/opt/local/lib" \
+	PKG_CONFIG_PATH=/c/msys64/mingw64/lib/pkgconfig \
+	$pybin -m pip install --user keyfinder
+
+	pypath="$($pybin -c 'import keyfinder; print(keyfinder.__file__)')"
+	for pyso in "${pypath%/*}"/*.so; do
+		[ -e "$pyso" ] || break
+		patchelf --set-rpath "${libpath%/*}" "$pyso" ||
+			echo "WARNING: patchelf failed (only fatal on musl-based distros)"
+	done
+	
+	mv "$pypath"{,.bak}
+	(
+		printf 'import ctypes\nctypes.cdll.LoadLibrary("%s")\n' "$libpath"
+		cat "$pypath.bak"
+	) >"$pypath"
+
+	echo
+	echo libkeyfinder successfully installed to the following locations:
+	echo "  $libpath"
+	echo "  $pypath"
+}
+
+
+have_beatroot() {
+	$pybin -c 'import vampyhost, sys; plugs = vampyhost.list_plugins(); sys.exit(0 if "beatroot-vamp:beatroot" in plugs else 1)'
+}
+
+
+install_vamp() {
+	# windows support:
+	#   use msys2 in mingw-w64 mode
+	#   pacman -S --needed mingw-w64-x86_64-{ffmpeg,python,python-pip,vamp-plugin-sdk}
+	
+	$pybin -m pip install --user vamp
+
+	have_beatroot || {
+		printf '\033[33mcould not find the vamp beatroot plugin, building from source\033[0m\n'
+		(dl_files yolo https://code.soundsoftware.ac.uk/attachments/download/885/beatroot-vamp-v1.0.tar.gz)
+		sha512sum -c <(
+			echo "1f444d1d58ccf565c0adfe99f1a1aa62789e19f5071e46857e2adfbc9d453037bc1c4dcb039b02c16240e9b97f444aaff3afb625c86aa2470233e711f55b6874  -"
+		) <beatroot-vamp-v1.0.tar.gz
+		tar -xf beatroot-vamp-v1.0.tar.gz 
+		cd beatroot-vamp-v1.0
+		make -f Makefile.linux -j4
+		# /home/ed/vamp /home/ed/.vamp /usr/local/lib/vamp
+		mkdir ~/vamp
+		cp -pv beatroot-vamp.* ~/vamp/
+	}
+	
+	have_beatroot &&
+		printf '\033[32mfound the vamp beatroot plugin, nice\033[0m\n' ||
+		printf '\033[31mWARNING: could not find the vamp beatroot plugin, please install it for optimal results\033[0m\n'
+}
+
+
+# not in use because it kinda segfaults, also no windows support
+install_soundtouch() {
+	gitlab_tarball https://gitlab.com/api/v4/projects/soundtouch%2Fsoundtouch/releases
+	
+	tar -xvf soundtouch-*
+	rm -- *.tar.gz
+	cd soundtouch-*
+	
+	# https://github.com/jrising/pysoundtouch
+	./bootstrap
+	./configure --enable-integer-samples CXXFLAGS="-fPIC" --prefix="$HOME/pe/soundtouch"
+	make -j$(nproc || echo 4)
+	make install
+	
+	CFLAGS=-I$HOME/pe/soundtouch/include/ \
+	LDFLAGS=-L$HOME/pe/soundtouch/lib \
+	$pybin -m pip install --user git+https://github.com/snowxmas/pysoundtouch.git
+	
+	pypath="$($pybin -c 'import importlib; print(importlib.util.find_spec("soundtouch").origin)')"
+	libpath="$(echo "$HOME/pe/soundtouch/lib/")"
+	patchelf --set-rpath "$libpath" "$pypath"
+
+	echo
+	echo soundtouch successfully installed to the following locations:
+	echo "  $libpath"
+	echo "  $pypath"
+}
+
+
+[ "$1" = keyfinder ] && { install_keyfinder; exit $?; }
+[ "$1" = soundtouch ] && { install_soundtouch; exit $?; }
+[ "$1" = vamp ] && { install_vamp; exit $?; }
+
+echo no args provided, installing keyfinder and vamp
+install_keyfinder
+install_vamp

bin/mtag/sleep.py

@@ -0,0 +1,8 @@
+#!/usr/bin/env python
+
+import time
+import random
+
+v = random.random() * 6
+time.sleep(v)
+print(f"{v:.2f}")

copyparty/__main__.py

@@ -12,16 +12,19 @@ import re
 import os
 import sys
 import time
+import signal
 import shutil
 import filecmp
 import locale
 import argparse
+import threading
+import traceback
 from textwrap import dedent
 
 from .__init__ import E, WINDOWS, VT100, PY2
 from .__version__ import S_VERSION, S_BUILD_DT, CODENAME
 from .svchub import SvcHub
-from .util import py_desc, align_tab
+from .util import py_desc, align_tab, IMPLICATIONS
 
 HAVE_SSL = True
 try:
@@ -164,6 +167,16 @@ def configure_ssl_ciphers(al):
         sys.exit(0)
 
 
+def sighandler(signal=None, frame=None):
+    msg = [""] * 5
+    for th in threading.enumerate():
+        msg.append(str(th))
+        msg.extend(traceback.format_stack(sys._current_frames()[th.ident]))
+
+    msg.append("\n")
+    print("\n".join(msg))
+
+
 def main():
     time.strptime("19970815", "%Y%m%d")  # python#7980
     if WINDOWS:
@@ -241,15 +254,17 @@ def main():
     ap.add_argument("-a", metavar="ACCT", type=str, action="append", help="add account")
     ap.add_argument("-v", metavar="VOL", type=str, action="append", help="add volume")
     ap.add_argument("-q", action="store_true", help="quiet")
+    ap.add_argument("--log-conn", action="store_true", help="print tcp-server msgs")
     ap.add_argument("-ed", action="store_true", help="enable ?dots")
     ap.add_argument("-emp", action="store_true", help="enable markdown plugins")
     ap.add_argument("-mcr", metavar="SEC", type=int, default=60, help="md-editor mod-chk rate")
     ap.add_argument("-nw", action="store_true", help="disable writes (benchmark)")
     ap.add_argument("-nih", action="store_true", help="no info hostname")
     ap.add_argument("-nid", action="store_true", help="no info disk-usage")
+    ap.add_argument("--no-zip", action="store_true", help="disable download as zip/tar")
     ap.add_argument("--no-sendfile", action="store_true", help="disable sendfile (for debugging)")
     ap.add_argument("--no-scandir", action="store_true", help="disable scandir (for debugging)")
-    ap.add_argument("--urlform", type=str, default="print,get", help="how to handle url-forms")
+    ap.add_argument("--urlform", metavar="MODE", type=str, default="print,get", help="how to handle url-forms")
     ap.add_argument("--salt", type=str, default="hunter2", help="up2k file-hash salt")
 
     ap2 = ap.add_argument_group('database options')
@@ -264,11 +279,13 @@ def main():
     ap2.add_argument("-mtm", metavar="M=t,t,t", action="append", type=str, help="add/replace metadata mapping")
     ap2.add_argument("-mte", metavar="M,M,M", type=str, help="tags to index/display (comma-sep.)",
         default="circle,album,.tn,artist,title,.bpm,key,.dur,.q")
+    ap2.add_argument("-mtp", metavar="M=[f,]bin", action="append", type=str, help="read tag M using bin")
+    ap2.add_argument("--srch-time", metavar="SEC", type=int, default=30, help="search deadline")
 
     ap2 = ap.add_argument_group('SSL/TLS options')
     ap2.add_argument("--http-only", action="store_true", help="disable ssl/tls")
     ap2.add_argument("--https-only", action="store_true", help="disable plaintext")
-    ap2.add_argument("--ssl-ver", type=str, help="ssl/tls versions to allow")
+    ap2.add_argument("--ssl-ver", metavar="LIST", type=str, help="ssl/tls versions to allow")
     ap2.add_argument("--ciphers", metavar="LIST", help="set allowed ciphers")
     ap2.add_argument("--ssl-dbg", action="store_true", help="dump some tls info")
     ap2.add_argument("--ssl-log", metavar="PATH", help="log master secrets")
@@ -277,13 +294,7 @@ def main():
     # fmt: on
 
     # propagate implications
-    for k1, k2 in [
-        ["e2dsa", "e2ds"],
-        ["e2ds", "e2d"],
-        ["e2tsr", "e2ts"],
-        ["e2ts", "e2t"],
-        ["e2t", "e2d"],
-    ]:
+    for k1, k2 in IMPLICATIONS:
         if getattr(al, k1):
             setattr(al, k2, True)
 
@@ -312,6 +323,8 @@ def main():
             + "  (if you crash with codec errors then that is why)"
         )
 
+    # signal.signal(signal.SIGINT, sighandler)
+
     SvcHub(al).run()
 
 

copyparty/__version__.py

@@ -1,8 +1,8 @@
 # coding: utf-8
 
-VERSION = (0, 9, 8)
-CODENAME = "the strongest music server"
-BUILD_DT = (2021, 3, 15)
+VERSION = (0, 10, 2)
+CODENAME = "zip it"
+BUILD_DT = (2021, 3, 27)
 
 S_VERSION = ".".join(map(str, VERSION))
 S_BUILD_DT = "{0:04d}-{1:02d}-{2:02d}".format(*BUILD_DT)

copyparty/authsrv.py

@@ -1,12 +1,14 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
-import os
 import re
+import os
+import sys
+import stat
 import threading
 
 from .__init__ import PY2, WINDOWS
-from .util import undot, Pebkac, fsdec, fsenc, statdir, nuprint
+from .util import IMPLICATIONS, undot, Pebkac, fsdec, fsenc, statdir, nuprint
 
 
 class VFS(object):
@@ -21,6 +23,14 @@ class VFS(object):
         self.nodes = {}  # child nodes
         self.all_vols = {vpath: self}  # flattened recursive
 
+    def __repr__(self):
+        return "VFS({})".format(
+            ", ".join(
+                "{}={!r}".format(k, self.__dict__[k])
+                for k in "realpath vpath uread uwrite flags".split()
+            )
+        )
+
     def _trk(self, vol):
         self.all_vols[vol.vpath] = vol
         return vol
@@ -44,6 +54,7 @@ class VFS(object):
                 self.uwrite,
                 self.flags,
             )
+            self._trk(vn)
             self.nodes[name] = vn
             return self._trk(vn.add(src, dst))
 
@@ -118,6 +129,73 @@ class VFS(object):
 
         return [abspath, real, virt_vis]
 
+    def walk(self, rel, rem, uname, dots, scandir, lstat=False):
+        """
+        recursively yields from ./rem;
+        rel is a unix-style user-defined vpath (not vfs-related)
+        """
+
+        fsroot, vfs_ls, vfs_virt = self.ls(rem, uname, scandir, lstat)
+        rfiles = [x for x in vfs_ls if not stat.S_ISDIR(x[1].st_mode)]
+        rdirs = [x for x in vfs_ls if stat.S_ISDIR(x[1].st_mode)]
+
+        rfiles.sort()
+        rdirs.sort()
+
+        yield rel, fsroot, rfiles, rdirs, vfs_virt
+
+        for rdir, _ in rdirs:
+            if not dots and rdir.startswith("."):
+                continue
+
+            wrel = (rel + "/" + rdir).lstrip("/")
+            wrem = (rem + "/" + rdir).lstrip("/")
+            for x in self.walk(wrel, wrem, uname, scandir, lstat):
+                yield x
+
+        for n, vfs in sorted(vfs_virt.items()):
+            if not dots and n.startswith("."):
+                continue
+
+            wrel = (rel + "/" + n).lstrip("/")
+            for x in vfs.walk(wrel, "", uname, scandir, lstat):
+                yield x
+
+    def zipgen(self, vrem, flt, uname, dots, scandir):
+        if flt:
+            flt = {k: True for k in flt}
+
+        for vpath, apath, files, rd, vd in self.walk("", vrem, uname, dots, scandir):
+            if flt:
+                files = [x for x in files if x[0] in flt]
+                rd = [x for x in rd if x[0] in flt]
+                vd = {x: y for x, y in vd.items() if x in flt}
+                flt = None
+
+            # print(repr([vpath, apath, [x[0] for x in files]]))
+            fnames = [n[0] for n in files]
+            vpaths = [vpath + "/" + n for n in fnames] if vpath else fnames
+            apaths = [os.path.join(apath, n) for n in fnames]
+            files = list(zip(vpaths, apaths, files))
+
+            if not dots:
+                # dotfile filtering based on vpath (intended visibility)
+                files = [x for x in files if "/." not in "/" + x[0]]
+
+                rm = [x for x in rd if x[0].startswith(".")]
+                for x in rm:
+                    rd.remove(x)
+
+                rm = [k for k in vd.keys() if k.startswith(".")]
+                for x in rm:
+                    del vd[x]
+
+            # up2k filetring based on actual abspath
+            files = [x for x in files if "{0}.hist{0}up2k.".format(os.sep) not in x[1]]
+
+            for f in [{"vp": v, "ap": a, "st": n[1]} for v, a, n in files]:
+                yield f
+
     def user_tree(self, uname, readable=False, writable=False):
         ret = []
         opt1 = readable and (uname in self.uread or "*" in self.uread)
@@ -200,16 +278,39 @@ class AuthSrv(object):
                 continue
 
             lvl, uname = ln.split(" ")
-            if lvl in "ra":
-                mread[vol_dst].append(uname)
-            if lvl in "wa":
-                mwrite[vol_dst].append(uname)
-            if lvl == "c":
-                cval = True
-                if "=" in uname:
-                    uname, cval = uname.split("=", 1)
+            self._read_vol_str(
+                lvl, uname, mread[vol_dst], mwrite[vol_dst], mflags[vol_dst]
+            )
 
-                mflags[vol_dst][uname] = cval
+    def _read_vol_str(self, lvl, uname, mr, mw, mf):
+        if lvl == "c":
+            cval = True
+            if "=" in uname:
+                uname, cval = uname.split("=", 1)
+
+            self._read_volflag(mf, uname, cval, False)
+            return
+
+        if uname == "":
+            uname = "*"
+
+        if lvl in "ra":
+            mr.append(uname)
+
+        if lvl in "wa":
+            mw.append(uname)
+
+    def _read_volflag(self, flags, name, value, is_list):
+        if name not in ["mtp"]:
+            flags[name] = value
+            return
+
+        if not is_list:
+            value = [value]
+        elif not value:
+            return
+
+        flags[name] = flags.get(name, []) + value
 
     def reload(self):
         """
@@ -232,7 +333,7 @@ class AuthSrv(object):
 
         if self.args.v:
             # list of src:dst:permset:permset:...
-            # permset is [rwa]username
+            # permset is [rwa]username or [c]flag
             for v_str in self.args.v:
                 m = self.re_vol.match(v_str)
                 if not m:
@@ -249,22 +350,7 @@ class AuthSrv(object):
 
                 perms = perms.split(":")
                 for (lvl, uname) in [[x[0], x[1:]] for x in perms]:
-                    if lvl == "c":
-                        cval = True
-                        if "=" in uname:
-                            uname, cval = uname.split("=", 1)
-
-                        mflags[dst][uname] = cval
-                        continue
-
-                    if uname == "":
-                        uname = "*"
-
-                    if lvl in "ra":
-                        mread[dst].append(uname)
-
-                    if lvl in "wa":
-                        mwrite[dst].append(uname)
+                    self._read_vol_str(lvl, uname, mread[dst], mwrite[dst], mflags[dst])
 
         if self.args.c:
             for cfg_fn in self.args.c:
@@ -310,6 +396,8 @@ class AuthSrv(object):
             )
             raise Exception("invalid config")
 
+        all_mte = {}
+        errors = False
         for vol in vfs.all_vols.values():
             if (self.args.e2ds and vol.uwrite) or self.args.e2dsa:
                 vol.flags["e2ds"] = True
@@ -321,10 +409,75 @@ class AuthSrv(object):
                 if getattr(self.args, k):
                     vol.flags[k] = True
 
+            for k1, k2 in IMPLICATIONS:
+                if k1 in vol.flags:
+                    vol.flags[k2] = True
+
             # default tag-list if unset
             if "mte" not in vol.flags:
                 vol.flags["mte"] = self.args.mte
 
+            # append parsers from argv to volume-flags
+            self._read_volflag(vol.flags, "mtp", self.args.mtp, True)
+
+            # d2d drops all database features for a volume
+            for grp, rm in [["d2d", "e2d"], ["d2t", "e2t"]]:
+                if not vol.flags.get(grp, False):
+                    continue
+
+                vol.flags["d2t"] = True
+                vol.flags = {k: v for k, v in vol.flags.items() if not k.startswith(rm)}
+
+            # mt* needs e2t so drop those too
+            for grp, rm in [["e2t", "mt"]]:
+                if vol.flags.get(grp, False):
+                    continue
+
+                vol.flags = {k: v for k, v in vol.flags.items() if not k.startswith(rm)}
+
+            # verify tags mentioned by -mt[mp] are used by -mte
+            local_mtp = {}
+            local_only_mtp = {}
+            for a in vol.flags.get("mtp", []) + vol.flags.get("mtm", []):
+                a = a.split("=")[0]
+                local_mtp[a] = True
+                local = True
+                for b in self.args.mtp or []:
+                    b = b.split("=")[0]
+                    if a == b:
+                        local = False
+
+                if local:
+                    local_only_mtp[a] = True
+
+            local_mte = {}
+            for a in vol.flags.get("mte", "").split(","):
+                local = True
+                all_mte[a] = True
+                local_mte[a] = True
+                for b in self.args.mte.split(","):
+                    if not a or not b:
+                        continue
+
+                    if a == b:
+                        local = False
+
+            for mtp in local_only_mtp.keys():
+                if mtp not in local_mte:
+                    m = 'volume "/{}" defines metadata tag "{}", but doesnt use it in "-mte" (or with "cmte" in its volume-flags)'
+                    self.log(m.format(vol.vpath, mtp), 1)
+                    errors = True
+
+        for mtp in self.args.mtp or []:
+            mtp = mtp.split("=")[0]
+            if mtp not in all_mte:
+                m = 'metadata tag "{}" is defined by "-mtm" or "-mtp", but is not used by "-mte" (or by any "cmte" volume-flag)'
+                self.log(m.format(mtp), 1)
+                errors = True
+
+        if errors:
+            sys.exit(1)
+
         try:
             v, _ = vfs.get("/", "*", False, True)
             if self.warn_anonwrite and os.getcwd() == v.realpath:

copyparty/broker_mpw.py

@@ -73,7 +73,9 @@ class MpWorker(object):
                 if PY2:
                     sck = pickle.loads(sck)  # nosec
 
-                self.log("%s %s" % addr, "|%sC-qpop" % ("-" * 4,), c="1;30")
+                if self.args.log_conn:
+                    self.log("%s %s" % addr, "|%sC-qpop" % ("-" * 4,), c="1;30")
+                
                 self.httpsrv.accept(sck, addr)
 
                 with self.mutex:

copyparty/broker_thr.py

@@ -28,7 +28,9 @@ class BrokerThr(object):
     def put(self, want_retval, dest, *args):
         if dest == "httpconn":
             sck, addr = args
-            self.log("%s %s" % addr, "|%sC-qpop" % ("-" * 4,), c="1;30")
+            if self.args.log_conn:
+                self.log("%s %s" % addr, "|%sC-qpop" % ("-" * 4,), c="1;30")
+
             self.httpsrv.accept(sck, addr)
 
         else:

copyparty/httpcli.py

@@ -7,6 +7,7 @@ import gzip
 import time
 import copy
 import json
+import string
 import socket
 import ctypes
 from datetime import datetime
@@ -14,6 +15,8 @@ import calendar
 
 from .__init__ import E, PY2, WINDOWS
 from .util import *  # noqa  # pylint: disable=unused-wildcard-import
+from .szip import StreamZip
+from .star import StreamTar
 
 if not PY2:
     unicode = str
@@ -45,13 +48,17 @@ class HttpCli(object):
         self.log_func(self.log_src, msg, c)
 
     def _check_nonfatal(self, ex):
-        return ex.code < 400 or ex.code == 404
+        return ex.code < 400 or ex.code in [404, 429]
 
     def _assert_safe_rem(self, rem):
         # sanity check to prevent any disasters
         if rem.startswith("/") or rem.startswith("../") or "/../" in rem:
             raise Exception("that was close")
 
+    def j2(self, name, **kwargs):
+        tpl = self.conn.hsrv.j2[name]
+        return tpl.render(**kwargs) if kwargs else tpl
+
     def run(self):
         """returns true if connection can be reused"""
         self.keepalive = False
@@ -154,7 +161,9 @@ class HttpCli(object):
             try:
                 # self.log("pebkac at httpcli.run #2: " + repr(ex))
                 self.keepalive = self._check_nonfatal(ex)
-                self.loud_reply("{}: {}".format(str(ex), self.vpath), status=ex.code)
+                self.log("{}\033[0m, {}".format(str(ex), self.vpath), 3)
+                msg = "<pre>{}\r\nURL: {}\r\n".format(str(ex), self.vpath)
+                self.reply(msg.encode("utf-8", "replace"), status=ex.code)
                 return self.keepalive
             except Pebkac:
                 return False
@@ -388,8 +397,30 @@ class HttpCli(object):
         if act == "tput":
             return self.handle_text_upload()
 
+        if act == "zip":
+            return self.handle_zip_post()
+
         raise Pebkac(422, 'invalid action "{}"'.format(act))
 
+    def handle_zip_post(self):
+        for k in ["zip", "tar"]:
+            v = self.uparam.get(k)
+            if v is not None:
+                break
+
+        if v is None:
+            raise Pebkac(422, "need zip or tar keyword")
+
+        vn, rem = self.auth.vfs.get(self.vpath, self.uname, True, False)
+        items = self.parser.require("files", 1024 * 1024)
+        if not items:
+            raise Pebkac(422, "need files list")
+
+        items = items.replace("\r", "").split("\n")
+        items = [unquotep(x) for x in items if items]
+
+        return self.tx_zip(k, v, vn, rem, items, self.args.ed)
+
     def handle_post_json(self):
         try:
             remains = int(self.headers["content-length"])
@@ -417,15 +448,18 @@ class HttpCli(object):
         if "srch" in self.uparam or "srch" in body:
             return self.handle_search(body)
 
-        # prefer this over undot; no reason to allow traversion
-        if "/" in body["name"]:
-            raise Pebkac(400, "folders verboten")
-
         # up2k-php compat
         for k in "chunkpit.php", "handshake.php":
             if self.vpath.endswith(k):
                 self.vpath = self.vpath[: -len(k)]
 
+        sub = None
+        name = undot(body["name"])
+        if "/" in name:
+            sub, name = name.rsplit("/", 1)
+            self.vpath = "/".join([self.vpath, sub]).strip("/")
+            body["name"] = name
+
         vfs, rem = self.conn.auth.vfs.get(self.vpath, self.uname, False, True)
 
         body["vtop"] = vfs.vpath
@@ -434,12 +468,22 @@ class HttpCli(object):
         body["addr"] = self.ip
         body["vcfg"] = vfs.flags
 
-        x = self.conn.hsrv.broker.put(True, "up2k.handle_json", body)
-        response = x.get()
-        response = json.dumps(response)
+        if sub:
+            try:
+                dst = os.path.join(vfs.realpath, rem)
+                os.makedirs(dst)
+            except:
+                if not os.path.isdir(dst):
+                    raise Pebkac(400, "some file got your folder name")
 
-        self.log(response)
-        self.reply(response.encode("utf-8"), mime="application/json")
+        x = self.conn.hsrv.broker.put(True, "up2k.handle_json", body)
+        ret = x.get()
+        if sub:
+            ret["name"] = "/".join([sub, ret["name"]])
+
+        ret = json.dumps(ret)
+        self.log(ret)
+        self.reply(ret.encode("utf-8"), mime="application/json")
         return True
 
     def handle_search(self, body):
@@ -450,19 +494,30 @@ class HttpCli(object):
 
         idx = self.conn.get_u2idx()
         t0 = time.time()
+        if idx.p_end:
+            penalty = 0.7
+            t_idle = t0 - idx.p_end
+            if idx.p_dur > 0.7 and t_idle < penalty:
+                m = "rate-limit ({:.1f} sec), cost {:.2f}, idle {:.2f}"
+                raise Pebkac(429, m.format(penalty, idx.p_dur, t_idle))
+
         if "srch" in body:
             # search by up2k hashlist
             vbody = copy.deepcopy(body)
             vbody["hash"] = len(vbody["hash"])
             self.log("qj: " + repr(vbody))
             hits = idx.fsearch(vols, body)
-            self.log("q#: {} ({:.2f}s)".format(repr(hits), time.time() - t0))
+            msg = repr(hits)
             taglist = []
         else:
             # search by query params
             self.log("qj: " + repr(body))
             hits, taglist = idx.search(vols, body)
-            self.log("q#: {} ({:.2f}s)".format(len(hits), time.time() - t0))
+            msg = len(hits)
+
+        idx.p_end = time.time()
+        idx.p_dur = idx.p_end - t0
+        self.log("q#: {} ({:.2f}s)".format(msg, idx.p_dur))
 
         order = []
         cfg = self.args.mte.split(",")
@@ -569,7 +624,7 @@ class HttpCli(object):
             pwd = "x"  # nosec
 
         h = {"Set-Cookie": "cppwd={}; Path=/; SameSite=Lax".format(pwd)}
-        html = self.conn.tpl_msg.render(h1=msg, h2='<a href="/">ack</a>', redir="/")
+        html = self.j2("msg", h1=msg, h2='<a href="/">ack</a>', redir="/")
         self.reply(html.encode("utf-8"), headers=h)
         return True
 
@@ -600,7 +655,8 @@ class HttpCli(object):
 
         vpath = "{}/{}".format(self.vpath, sanitized).lstrip("/")
         esc_paths = [quotep(vpath), html_escape(vpath)]
-        html = self.conn.tpl_msg.render(
+        html = self.j2(
+            "msg",
             h2='<a href="/{}">go to /{}</a>'.format(*esc_paths),
             pre="aight",
             click=True,
@@ -632,7 +688,8 @@ class HttpCli(object):
                 f.write(b"`GRUNNUR`\n")
 
         vpath = "{}/{}".format(self.vpath, sanitized).lstrip("/")
-        html = self.conn.tpl_msg.render(
+        html = self.j2(
+            "msg",
             h2='<a href="/{}?edit">go to /{}?edit</a>'.format(
                 quotep(vpath), html_escape(vpath)
             ),
@@ -738,7 +795,8 @@ class HttpCli(object):
                     ).encode("utf-8")
                 )
 
-        html = self.conn.tpl_msg.render(
+        html = self.j2(
+            "msg",
             h2='<a href="/{}">return to /{}</a>'.format(
                 quotep(self.vpath), html_escape(self.vpath)
             ),
@@ -1026,16 +1084,75 @@ class HttpCli(object):
         self.log("{},  {}".format(logmsg, spd))
         return ret
 
+    def tx_zip(self, fmt, uarg, vn, rem, items, dots):
+        if self.args.no_zip:
+            raise Pebkac(400, "not enabled")
+
+        logmsg = "{:4} {} ".format("", self.req)
+        self.keepalive = False
+
+        if not uarg:
+            uarg = ""
+
+        if fmt == "tar":
+            mime = "application/x-tar"
+            packer = StreamTar
+        else:
+            mime = "application/zip"
+            packer = StreamZip
+
+        fn = items[0] if items and items[0] else self.vpath
+        if fn:
+            fn = fn.rstrip("/").split("/")[-1]
+        else:
+            fn = self.headers.get("host", "hey")
+
+        afn = "".join(
+            [x if x in (string.ascii_letters + string.digits) else "_" for x in fn]
+        )
+
+        bascii = unicode(string.ascii_letters + string.digits).encode("utf-8")
+        ufn = fn.encode("utf-8", "xmlcharrefreplace")
+        if PY2:
+            ufn = [unicode(x) if x in bascii else "%{:02x}".format(ord(x)) for x in ufn]
+        else:
+            ufn = [
+                chr(x).encode("utf-8")
+                if x in bascii
+                else "%{:02x}".format(x).encode("ascii")
+                for x in ufn
+            ]
+        ufn = b"".join(ufn).decode("ascii")
+
+        cdis = "attachment; filename=\"{}.{}\"; filename*=UTF-8''{}.{}"
+        cdis = cdis.format(afn, fmt, ufn, fmt)
+        self.send_headers(None, mime=mime, headers={"Content-Disposition": cdis})
+
+        fgen = vn.zipgen(rem, items, self.uname, dots, not self.args.no_scandir)
+        # for f in fgen: print(repr({k: f[k] for k in ["vp", "ap"]}))
+        bgen = packer(fgen, utf8="utf" in uarg, pre_crc="crc" in uarg)
+        bsent = 0
+        for buf in bgen.gen():
+            if not buf:
+                break
+
+            try:
+                self.s.sendall(buf)
+                bsent += len(buf)
+            except:
+                logmsg += " \033[31m" + unicode(bsent) + "\033[0m"
+                break
+
+        spd = self._spd(bsent)
+        self.log("{},  {}".format(logmsg, spd))
+        return True
+
     def tx_md(self, fs_path):
         logmsg = "{:4} {} ".format("", self.req)
-        if "edit2" in self.uparam:
-            html_path = "web/mde.html"
-            template = self.conn.tpl_mde
-        else:
-            html_path = "web/md.html"
-            template = self.conn.tpl_md
 
-        html_path = os.path.join(E.mod, html_path)
+        tpl = "mde" if "edit2" in self.uparam else "md"
+        html_path = os.path.join(E.mod, "web", "{}.html".format(tpl))
+        template = self.j2(tpl)
 
         st = os.stat(fsenc(fs_path))
         # sz_md = st.st_size
@@ -1087,7 +1204,7 @@ class HttpCli(object):
     def tx_mounts(self):
         rvol = [x + "/" if x else x for x in self.rvol]
         wvol = [x + "/" if x else x for x in self.wvol]
-        html = self.conn.tpl_mounts.render(this=self, rvol=rvol, wvol=wvol)
+        html = self.j2("splash", this=self, rvol=rvol, wvol=wvol)
         self.reply(html.encode("utf-8"))
         return True
 
@@ -1176,6 +1293,11 @@ class HttpCli(object):
 
             return self.tx_file(abspath)
 
+        for k in ["zip", "tar"]:
+            v = self.uparam.get(k)
+            if v is not None:
+                return self.tx_zip(k, v, vn, rem, [], self.args.ed)
+
         fsroot, vfs_ls, vfs_virt = vn.ls(rem, self.uname, not self.args.no_scandir)
         stats = {k: v for k, v in vfs_ls}
         vfs_ls = [x[0] for x in vfs_ls]
@@ -1236,8 +1358,11 @@ class HttpCli(object):
 
             is_dir = stat.S_ISDIR(inf.st_mode)
             if is_dir:
-                margin = "DIR"
                 href += "/"
+                if self.args.no_zip:
+                    margin = "DIR"
+                else:
+                    margin = '<a href="{}?zip">zip</a>'.format(quotep(href))
             elif fn in hist:
                 margin = '<a href="{}.hist/{}">#{}</a>'.format(
                     base, html_escape(hist[fn][2], quote=True), hist[fn][0]
@@ -1284,7 +1409,7 @@ class HttpCli(object):
 
                 tags = {}
                 f["tags"] = tags
-                
+
                 if not r:
                     continue
 
@@ -1295,7 +1420,7 @@ class HttpCli(object):
                     tags[k] = v
 
         if icur:
-            taglist = [k for k in self.args.mte.split(",") if k in taglist]
+            taglist = [k for k in vn.flags.get("mte", "").split(",") if k in taglist]
             for f in dirs:
                 f["tags"] = {}
 
@@ -1361,16 +1486,20 @@ class HttpCli(object):
 
         dirs.extend(files)
 
-        html = self.conn.tpl_browser.render(
+        html = self.j2(
+            "browser",
             vdir=quotep(self.vpath),
             vpnodes=vpnodes,
             files=dirs,
             ts=ts,
             perms=json.dumps(perms),
             taglist=taglist,
-            tag_order=json.dumps(self.args.mte.split(",")),
+            tag_order=json.dumps(
+                vn.flags["mte"].split(",") if "mte" in vn.flags else []
+            ),
             have_up2k_idx=("e2d" in vn.flags),
             have_tags_idx=("e2t" in vn.flags),
+            have_zip=(not self.args.no_zip),
             logues=logues,
             title=html_escape(self.vpath),
             srv_info=srv_info,

copyparty/httpconn.py

@@ -12,23 +12,6 @@ try:
 except:
     HAVE_SSL = False
 
-try:
-    import jinja2
-except ImportError:
-    print(
-        """\033[1;31m
-  you do not have jinja2 installed,\033[33m
-  choose one of these:\033[0m
-   * apt install python-jinja2
-   * {} -m pip install --user jinja2
-   * (try another python version, if you have one)
-   * (try copyparty.sfx instead)
-""".format(
-            os.path.basename(sys.executable)
-        )
-    )
-    sys.exit(1)
-
 from .__init__ import E
 from .util import Unrecv
 from .httpcli import HttpCli
@@ -57,14 +40,6 @@ class HttpConn(object):
         self.log_func = hsrv.log
         self.set_rproxy()
 
-        env = jinja2.Environment()
-        env.loader = jinja2.FileSystemLoader(os.path.join(E.mod, "web"))
-        self.tpl_mounts = env.get_template("splash.html")
-        self.tpl_browser = env.get_template("browser.html")
-        self.tpl_msg = env.get_template("msg.html")
-        self.tpl_md = env.get_template("md.html")
-        self.tpl_mde = env.get_template("mde.html")
-
     def set_rproxy(self, ip=None):
         if ip is None:
             color = 36
@@ -112,7 +87,9 @@ class HttpConn(object):
                 err = "need at least 4 bytes in the first packet; got {}".format(
                     len(method)
                 )
-                self.log(err)
+                if method:
+                    self.log(err)
+
                 self.s.send(b"HTTP/1.1 400 Bad Request\r\n\r\n" + err.encode("utf-8"))
                 return
 

copyparty/httpsrv.py

@@ -2,10 +2,28 @@
 from __future__ import print_function, unicode_literals
 
 import os
+import sys
 import time
 import socket
 import threading
 
+try:
+    import jinja2
+except ImportError:
+    print(
+        """\033[1;31m
+  you do not have jinja2 installed,\033[33m
+  choose one of these:\033[0m
+   * apt install python-jinja2
+   * {} -m pip install --user jinja2
+   * (try another python version, if you have one)
+   * (try copyparty.sfx instead)
+""".format(
+            os.path.basename(sys.executable)
+        )
+    )
+    sys.exit(1)
+
 from .__init__ import E, MACOS
 from .httpconn import HttpConn
 from .authsrv import AuthSrv
@@ -30,6 +48,13 @@ class HttpSrv(object):
         self.workload_thr_alive = False
         self.auth = AuthSrv(self.args, self.log)
 
+        env = jinja2.Environment()
+        env.loader = jinja2.FileSystemLoader(os.path.join(E.mod, "web"))
+        self.j2 = {
+            x: env.get_template(x + ".html")
+            for x in ["splash", "browser", "msg", "md", "mde"]
+        }
+
         cert_path = os.path.join(E.cfg, "cert.pem")
         if os.path.exists(cert_path):
             self.cert_path = cert_path
@@ -38,7 +63,9 @@ class HttpSrv(object):
 
     def accept(self, sck, addr):
         """takes an incoming tcp connection and creates a thread to handle it"""
-        self.log("%s %s" % addr, "|%sC-cthr" % ("-" * 5,), c="1;30")
+        if self.args.log_conn:
+            self.log("%s %s" % addr, "|%sC-cthr" % ("-" * 5,), c="1;30")
+
         thr = threading.Thread(target=self.thr_client, args=(sck, addr))
         thr.daemon = True
         thr.start()
@@ -66,11 +93,15 @@ class HttpSrv(object):
                 thr.start()
 
         try:
-            self.log("%s %s" % addr, "|%sC-crun" % ("-" * 6,), c="1;30")
+            if self.args.log_conn:
+                self.log("%s %s" % addr, "|%sC-crun" % ("-" * 6,), c="1;30")
+
             cli.run()
 
         finally:
-            self.log("%s %s" % addr, "|%sC-cdone" % ("-" * 7,), c="1;30")
+            if self.args.log_conn:
+                self.log("%s %s" % addr, "|%sC-cdone" % ("-" * 7,), c="1;30")
+
             try:
                 sck.shutdown(socket.SHUT_RDWR)
                 sck.close()

copyparty/mtag.py

@@ -8,7 +8,7 @@ import shutil
 import subprocess as sp
 
 from .__init__ import PY2, WINDOWS
-from .util import fsenc, fsdec
+from .util import fsenc, fsdec, REKOBO_LKEY
 
 if not PY2:
     unicode = str
@@ -151,6 +151,12 @@ class MTag(object):
                 v = v.split("/")[0].strip().lstrip("0")
                 ret[k] = v or 0
 
+        # normalize key notation to rkeobo
+        okey = ret.get("key")
+        if okey:
+            key = okey.replace(" ", "").replace("maj", "").replace("min", "m")
+            ret["key"] = REKOBO_LKEY.get(key.lower(), okey)
+
         return ret
 
     def compare(self, abspath):
@@ -225,7 +231,7 @@ class MTag(object):
         """
         note:
           tags which contain newline will be truncated on first \n,
-          ffmpeg emits \n and spacepads the : to align visually
+          ffprobe emits \n and spacepads the : to align visually
         note:
           the Stream ln always mentions Audio: if audio
           the Stream ln usually has kb/s, is more accurate
@@ -295,7 +301,7 @@ class MTag(object):
                             sec *= 60
                             sec += int(f)
                     except:
-                        self.log("invalid timestr from ffmpeg: [{}]".format(tstr), c=3)
+                        self.log("invalid timestr from ffprobe: [{}]".format(tstr), c=3)
 
                 ret[".dur"] = sec
                 m = ptn_br1.search(ln)
@@ -312,3 +318,30 @@ class MTag(object):
         ret = {k: [0, v] for k, v in ret.items()}
 
         return self.normalize_tags(ret, md)
+
+    def get_bin(self, parsers, abspath):
+        pypath = os.path.abspath(os.path.dirname(os.path.dirname(__file__)))
+        pypath = [str(pypath)] + [str(x) for x in sys.path if x]
+        pypath = str(os.pathsep.join(pypath))
+        env = os.environ.copy()
+        env["PYTHONPATH"] = pypath
+
+        ret = {}
+        for tagname, (binpath, timeout) in parsers.items():
+            try:
+                cmd = [sys.executable, binpath, abspath]
+                args = {"env": env, "timeout": timeout}
+
+                if WINDOWS:
+                    args["creationflags"] = 0x4000
+                else:
+                    cmd = ["nice"] + cmd
+
+                cmd = [fsenc(x) for x in cmd]
+                v = sp.check_output(cmd, **args).strip()
+                if v:
+                    ret[tagname] = v.decode("utf-8")
+            except:
+                pass
+
+        return ret

copyparty/star.py

@@ -0,0 +1,84 @@
+import os
+import tarfile
+import threading
+
+from .sutil import errdesc
+from .util import Queue, fsenc
+
+
+class QFile(object):
+    """file-like object which buffers writes into a queue"""
+
+    def __init__(self):
+        self.q = Queue(64)
+
+    def write(self, buf):
+        self.q.put(buf)
+
+
+class StreamTar(object):
+    """construct in-memory tar file from the given path"""
+
+    def __init__(self, fgen, **kwargs):
+        self.ci = 0
+        self.co = 0
+        self.qfile = QFile()
+        self.fgen = fgen
+        self.errf = None
+
+        # python 3.8 changed to PAX_FORMAT as default,
+        # waste of space and don't care about the new features
+        fmt = tarfile.GNU_FORMAT
+        self.tar = tarfile.open(fileobj=self.qfile, mode="w|", format=fmt)
+
+        w = threading.Thread(target=self._gen)
+        w.daemon = True
+        w.start()
+
+    def gen(self):
+        while True:
+            buf = self.qfile.q.get()
+            if buf is None:
+                break
+
+            self.co += len(buf)
+            yield buf
+
+        yield None
+        if self.errf:
+            os.unlink(self.errf["ap"])
+
+    def ser(self, f):
+        name = f["vp"]
+        src = f["ap"]
+        fsi = f["st"]
+
+        inf = tarfile.TarInfo(name=name)
+        inf.mode = fsi.st_mode
+        inf.size = fsi.st_size
+        inf.mtime = fsi.st_mtime
+        inf.uid = 0
+        inf.gid = 0
+
+        self.ci += inf.size
+        with open(fsenc(src), "rb", 512 * 1024) as f:
+            self.tar.addfile(inf, f)
+
+    def _gen(self):
+        errors = []
+        for f in self.fgen:
+            if "err" in f:
+                errors.append([f["vp"], f["err"]])
+                continue
+
+            try:
+                self.ser(f)
+            except Exception as ex:
+                errors.append([f["vp"], repr(ex)])
+
+        if errors:
+            self.errf = errdesc(errors)
+            self.ser(self.errf)
+
+        self.tar.close()
+        self.qfile.q.put(None)

copyparty/sutil.py

@@ -0,0 +1,25 @@
+import os
+import time
+import tempfile
+from datetime import datetime
+
+
+def errdesc(errors):
+    report = ["copyparty failed to add the following files to the archive:", ""]
+
+    for fn, err in errors:
+        report.extend([" file: {}".format(fn), "error: {}".format(err), ""])
+
+    with tempfile.NamedTemporaryFile(prefix="copyparty-", delete=False) as tf:
+        tf_path = tf.name
+        tf.write("\r\n".join(report).encode("utf-8", "replace"))
+
+    dt = datetime.utcfromtimestamp(time.time())
+    dt = dt.strftime("%Y-%m%d-%H%M%S")
+
+    os.chmod(tf_path, 0o444)
+    return {
+        "vp": "archive-errors-{}.txt".format(dt),
+        "ap": tf_path,
+        "st": os.stat(tf_path),
+    }

copyparty/szip.py

@@ -0,0 +1,271 @@
+import os
+import time
+import zlib
+import struct
+from datetime import datetime
+
+from .sutil import errdesc
+from .util import yieldfile, sanitize_fn
+
+
+def dostime2unix(buf):
+    t, d = struct.unpack("<HH", buf)
+
+    ts = (t & 0x1F) * 2
+    tm = (t >> 5) & 0x3F
+    th = t >> 11
+
+    dd = d & 0x1F
+    dm = (d >> 5) & 0xF
+    dy = (d >> 9) + 1980
+
+    tt = (dy, dm, dd, th, tm, ts)
+    tf = "{:04d}-{:02d}-{:02d} {:02d}:{:02d}:{:02d}"
+    iso = tf.format(*tt)
+
+    dt = datetime.strptime(iso, "%Y-%m-%d %H:%M:%S")
+    return int(dt.timestamp())
+
+
+def unixtime2dos(ts):
+    tt = time.gmtime(ts)
+    dy, dm, dd, th, tm, ts = list(tt)[:6]
+
+    bd = ((dy - 1980) << 9) + (dm << 5) + dd
+    bt = (th << 11) + (tm << 5) + ts // 2
+    return struct.pack("<HH", bt, bd)
+
+
+def gen_fdesc(sz, crc32, z64):
+    ret = b"\x50\x4b\x07\x08"
+    fmt = "<LQQ" if z64 else "<LLL"
+    ret += struct.pack(fmt, crc32, sz, sz)
+    return ret
+
+
+def gen_hdr(h_pos, fn, sz, lastmod, utf8, crc32, pre_crc):
+    """
+    does regular file headers
+    and the central directory meme if h_pos is set
+    (h_pos = absolute position of the regular header)
+    """
+
+    # appnote 4.5 / zip 3.0 (2008) / unzip 6.0 (2009) says to add z64
+    # extinfo for values which exceed H, but that becomes an off-by-one
+    # (can't tell if it was clamped or exactly maxval), make it obvious
+    z64 = sz >= 0xFFFFFFFF
+    z64v = [sz, sz] if z64 else []
+    if h_pos and h_pos >= 0xFFFFFFFF:
+        # central, also consider ptr to original header
+        z64v.append(h_pos)
+
+    # confusingly this doesn't bump if h_pos
+    req_ver = b"\x2d\x00" if z64 else b"\x0a\x00"
+
+    if crc32:
+        crc32 = struct.pack("<L", crc32)
+    else:
+        crc32 = b"\x00" * 4
+
+    if h_pos is None:
+        # 4b magic, 2b min-ver
+        ret = b"\x50\x4b\x03\x04" + req_ver
+    else:
+        # 4b magic, 2b spec-ver, 2b min-ver
+        ret = b"\x50\x4b\x01\x02\x1e\x03" + req_ver
+
+    ret += b"\x00" if pre_crc else b"\x08"  # streaming
+    ret += b"\x08" if utf8 else b"\x00"  # appnote 6.3.2 (2007)
+
+    # 2b compression, 4b time, 4b crc
+    ret += b"\x00\x00" + unixtime2dos(lastmod) + crc32
+
+    # spec says to put zeros when !crc if bit3 (streaming)
+    # however infozip does actual sz and it even works on winxp
+    # (same reasning for z64 extradata later)
+    vsz = 0xFFFFFFFF if z64 else sz
+    ret += struct.pack("<LL", vsz, vsz)
+
+    # windows support (the "?" replace below too)
+    fn = sanitize_fn(fn, "/")
+    bfn = fn.encode("utf-8" if utf8 else "cp437", "replace").replace(b"?", b"_")
+
+    z64_len = len(z64v) * 8 + 4 if z64v else 0
+    ret += struct.pack("<HH", len(bfn), z64_len)
+
+    if h_pos is not None:
+        # 2b comment, 2b diskno
+        ret += b"\x00" * 4
+
+        # 2b internal.attr, 4b external.attr
+        # infozip-macos: 0100 0000 a481 file:644
+        # infozip-macos: 0100 0100 0080 file:000
+        ret += b"\x01\x00\x00\x00\xa4\x81"
+
+        # 4b local-header-ofs
+        ret += struct.pack("<L", min(h_pos, 0xFFFFFFFF))
+
+    ret += bfn
+
+    if z64v:
+        ret += struct.pack("<HH" + "Q" * len(z64v), 1, len(z64v) * 8, *z64v)
+
+    return ret
+
+
+def gen_ecdr(items, cdir_pos, cdir_end):
+    """
+    summary of all file headers,
+    usually the zipfile footer unless something clamps
+    """
+
+    ret = b"\x50\x4b\x05\x06"
+
+    # 2b ndisk, 2b disk0
+    ret += b"\x00" * 4
+
+    cdir_sz = cdir_end - cdir_pos
+
+    nitems = min(0xFFFF, len(items))
+    csz = min(0xFFFFFFFF, cdir_sz)
+    cpos = min(0xFFFFFFFF, cdir_pos)
+
+    need_64 = nitems == 0xFFFF or 0xFFFFFFFF in [csz, cpos]
+
+    # 2b tnfiles, 2b dnfiles, 4b dir sz, 4b dir pos
+    ret += struct.pack("<HHLL", nitems, nitems, csz, cpos)
+
+    # 2b comment length
+    ret += b"\x00\x00"
+
+    return [ret, need_64]
+
+
+def gen_ecdr64(items, cdir_pos, cdir_end):
+    """
+    z64 end of central directory
+    added when numfiles or a headerptr clamps
+    """
+
+    ret = b"\x50\x4b\x06\x06"
+
+    # 8b own length from hereon
+    ret += b"\x2c" + b"\x00" * 7
+
+    # 2b spec-ver, 2b min-ver
+    ret += b"\x1e\x03\x2d\x00"
+
+    # 4b ndisk, 4b disk0
+    ret += b"\x00" * 8
+
+    # 8b tnfiles, 8b dnfiles, 8b dir sz, 8b dir pos
+    cdir_sz = cdir_end - cdir_pos
+    ret += struct.pack("<QQQQ", len(items), len(items), cdir_sz, cdir_pos)
+
+    return ret
+
+
+def gen_ecdr64_loc(ecdr64_pos):
+    """
+    z64 end of central directory locator
+    points to ecdr64
+    why
+    """
+
+    ret = b"\x50\x4b\x06\x07"
+
+    # 4b cdisk, 8b start of ecdr64, 4b ndisks
+    ret += struct.pack("<LQL", 0, ecdr64_pos, 1)
+
+    return ret
+
+
+class StreamZip(object):
+    def __init__(self, fgen, utf8=False, pre_crc=False):
+        self.fgen = fgen
+        self.utf8 = utf8
+        self.pre_crc = pre_crc
+
+        self.pos = 0
+        self.items = []
+
+    def _ct(self, buf):
+        self.pos += len(buf)
+        return buf
+
+    def ser(self, f):
+        name = f["vp"]
+        src = f["ap"]
+        st = f["st"]
+
+        sz = st.st_size
+        ts = st.st_mtime + 1
+
+        crc = None
+        if self.pre_crc:
+            crc = 0
+            for buf in yieldfile(src):
+                crc = zlib.crc32(buf, crc)
+
+            crc &= 0xFFFFFFFF
+
+        h_pos = self.pos
+        buf = gen_hdr(None, name, sz, ts, self.utf8, crc, self.pre_crc)
+        yield self._ct(buf)
+
+        crc = crc or 0
+        for buf in yieldfile(src):
+            if not self.pre_crc:
+                crc = zlib.crc32(buf, crc)
+
+            yield self._ct(buf)
+
+        crc &= 0xFFFFFFFF
+
+        self.items.append([name, sz, ts, crc, h_pos])
+
+        z64 = sz >= 4 * 1024 * 1024 * 1024
+
+        if z64 or not self.pre_crc:
+            buf = gen_fdesc(sz, crc, z64)
+            yield self._ct(buf)
+
+    def gen(self):
+        errors = []
+        for f in self.fgen:
+            if "err" in f:
+                errors.append([f["vp"], f["err"]])
+                continue
+
+            try:
+                for x in self.ser(f):
+                    yield x
+            except Exception as ex:
+                errors.append([f["vp"], repr(ex)])
+
+        if errors:
+            errf = errdesc(errors)
+            print(repr(errf))
+            for x in self.ser(errf):
+                yield x
+
+        cdir_pos = self.pos
+        for name, sz, ts, crc, h_pos in self.items:
+            buf = gen_hdr(h_pos, name, sz, ts, self.utf8, crc, self.pre_crc)
+            yield self._ct(buf)
+        cdir_end = self.pos
+
+        _, need_64 = gen_ecdr(self.items, cdir_pos, cdir_end)
+        if need_64:
+            ecdir64_pos = self.pos
+            buf = gen_ecdr64(self.items, cdir_pos, cdir_end)
+            yield self._ct(buf)
+
+            buf = gen_ecdr64_loc(ecdir64_pos)
+            yield self._ct(buf)
+
+        ecdr, _ = gen_ecdr(self.items, cdir_pos, cdir_end)
+        yield self._ct(ecdr)
+
+        if errors:
+            os.unlink(errf["ap"])

copyparty/tcpsrv.py

@@ -68,23 +68,29 @@ class TcpSrv(object):
             self.log("tcpsrv", "listening @ {0}:{1}".format(ip, port))
 
         while True:
-            self.log("tcpsrv", "|%sC-ncli" % ("-" * 1,), c="1;30")
+            if self.args.log_conn:
+                self.log("tcpsrv", "|%sC-ncli" % ("-" * 1,), c="1;30")
+
             if self.num_clients.v >= self.args.nc:
                 time.sleep(0.1)
                 continue
 
-            self.log("tcpsrv", "|%sC-acc1" % ("-" * 2,), c="1;30")
+            if self.args.log_conn:
+                self.log("tcpsrv", "|%sC-acc1" % ("-" * 2,), c="1;30")
+
             ready, _, _ = select.select(self.srv, [], [])
             for srv in ready:
                 sck, addr = srv.accept()
                 sip, sport = srv.getsockname()
-                self.log(
-                    "%s %s" % addr,
-                    "|{}C-acc2 \033[0;36m{} \033[3{}m{}".format(
-                        "-" * 3, sip, sport % 8, sport
-                    ),
-                    c="1;30",
-                )
+                if self.args.log_conn:
+                    self.log(
+                        "%s %s" % addr,
+                        "|{}C-acc2 \033[0;36m{} \033[3{}m{}".format(
+                            "-" * 3, sip, sport % 8, sport
+                        ),
+                        c="1;30",
+                    )
+
                 self.num_clients.add()
                 self.hub.broker.put(False, "httpconn", sck, addr)
 

copyparty/u2idx.py

@@ -1,10 +1,13 @@
 # coding: utf-8
 from __future__ import print_function, unicode_literals
 
+import re
 import os
+import time
+import threading
 from datetime import datetime
 
-from .util import u8safe
+from .util import u8safe, s3dec, html_escape, Pebkac
 from .up2k import up2k_wark_from_hashlist
 
 
@@ -19,6 +22,7 @@ class U2idx(object):
     def __init__(self, args, log_func):
         self.args = args
         self.log_func = log_func
+        self.timeout = args.srch_time
 
         if not HAVE_SQLITE3:
             self.log("could not load sqlite3; searchign wqill be disabled")
@@ -28,6 +32,9 @@ class U2idx(object):
         self.mem_cur = sqlite3.connect(":memory:")
         self.mem_cur.execute(r"create table a (b text)")
 
+        self.p_end = None
+        self.p_dur = 0
+
     def log(self, msg, c=0):
         self.log_func("u2idx", msg, c)
 
@@ -43,7 +50,10 @@ class U2idx(object):
         uq = "substr(w,1,16) = ? and w = ?"
         uv = [wark[:16], wark]
 
-        return self.run_query(vols, uq, uv, "", [])[0]
+        try:
+            return self.run_query(vols, uq, uv, {})[0]
+        except Exception as ex:
+            raise Pebkac(500, repr(ex))
 
     def get_cur(self, ptop):
         cur = self.cur.get(ptop)
@@ -73,17 +83,64 @@ class U2idx(object):
 
         uq, uv = _sqlize(qobj)
 
-        tq = ""
-        tv = []
         qobj = {}
         if "tags" in body:
             _conv_txt(qobj, body, "tags", "mt.v")
-            tq, tv = _sqlize(qobj)
 
-        return self.run_query(vols, uq, uv, tq, tv)
+        if "adv" in body:
+            _conv_adv(qobj, body, "adv")
 
-    def run_query(self, vols, uq, uv, tq, tv):
-        self.log("qs: {} {} ,  {} {}".format(uq, repr(uv), tq, repr(tv)))
+        try:
+            return self.run_query(vols, uq, uv, qobj)
+        except Exception as ex:
+            raise Pebkac(500, repr(ex))
+
+    def run_query(self, vols, uq, uv, targs):
+        self.log("qs: {} {} ,  {}".format(uq, repr(uv), repr(targs)))
+
+        done_flag = []
+        self.active_id = "{:.6f}_{}".format(
+            time.time(), threading.current_thread().ident
+        )
+        thr = threading.Thread(
+            target=self.terminator,
+            args=(
+                self.active_id,
+                done_flag,
+            ),
+        )
+        thr.daemon = True
+        thr.start()
+
+        if not targs:
+            if not uq:
+                q = "select * from up"
+                v = ()
+            else:
+                q = "select * from up where " + uq
+                v = tuple(uv)
+        else:
+            q = "select up.* from up"
+            keycmp = "substr(up.w,1,16)"
+            where = []
+            v = []
+            ctr = 0
+            for tq, tv in sorted(targs.items()):
+                ctr += 1
+                tq = tq.split("\n")[0]
+                keycmp2 = "mt{}.w".format(ctr)
+                q += " inner join mt mt{} on {} = {}".format(ctr, keycmp, keycmp2)
+                keycmp = keycmp2
+                where.append(tq.replace("mt.", keycmp[:-1]))
+                v.append(tv)
+
+            if uq:
+                where.append(uq)
+                v.extend(uv)
+
+            q += " where " + (" and ".join(where))
+
+        # self.log("q2: {} {}".format(q, repr(v)))
 
         ret = []
         lim = 1000
@@ -93,18 +150,7 @@ class U2idx(object):
             if not cur:
                 continue
 
-            if not tq:
-                if not uq:
-                    q = "select * from up"
-                    v = ()
-                else:
-                    q = "select * from up where " + uq
-                    v = tuple(uv)
-            else:
-                # naive assumption: tags first
-                q = "select up.* from up inner join mt on substr(up.w,1,16) = mt.w where {}"
-                q = q.format(" and ".join([tq, uq]) if uq else tq)
-                v = tuple(tv + uv)
+            self.active_cur = cur
 
             sret = []
             c = cur.execute(q, v)
@@ -124,17 +170,35 @@ class U2idx(object):
                 w = hit["w"]
                 del hit["w"]
                 tags = {}
-                q = "select k, v from mt where w = ? and k != 'x'"
-                for k, v in cur.execute(q, (w,)):
+                q2 = "select k, v from mt where w = ? and k != 'x'"
+                for k, v2 in cur.execute(q2, (w,)):
                     taglist[k] = True
-                    tags[k] = v
+                    tags[k] = v2
 
                 hit["tags"] = tags
 
             ret.extend(sret)
 
+        done_flag.append(True)
+        self.active_id = None
+
+        # undupe hits from multiple metadata keys
+        if len(ret) > 1:
+            ret = [ret[0]] + [
+                y for x, y in zip(ret[:-1], ret[1:]) if x["rp"] != y["rp"]
+            ]
+
         return ret, list(taglist.keys())
 
+    def terminator(self, identifier, done_flag):
+        for _ in range(self.timeout):
+            time.sleep(1)
+            if done_flag:
+                return
+
+        if identifier == self.active_id:
+            self.active_cur.connection.interrupt()
+
 
 def _open(ptop):
     db_path = os.path.join(ptop, ".hist", "up2k.db")
@@ -190,6 +254,23 @@ def _conv_txt(q, body, k, sql):
         q[qk + "\n" + v] = u8safe(v)
 
 
+def _conv_adv(q, body, k):
+    ptn = re.compile(r"^(\.?[a-z]+) *(==?|!=|<=?|>=?) *(.*)$")
+
+    parts = body[k].split(" ")
+    parts = [x.strip() for x in parts if x.strip()]
+
+    for part in parts:
+        m = ptn.match(part)
+        if not m:
+            p = html_escape(part)
+            raise Pebkac(400, "invalid argument [" + p + "]")
+
+        k, op, v = m.groups()
+        qk = "mt.k = '{}' and mt.v {} ?".format(k, op)
+        q[qk + "\n" + v] = u8safe(v)
+
+
 def _sqlize(qobj):
     keys = []
     values = []

copyparty/up2k.py

@@ -13,6 +13,7 @@ import base64
 import hashlib
 import threading
 import traceback
+import subprocess as sp
 from copy import deepcopy
 
 from .__init__ import WINDOWS
@@ -28,6 +29,7 @@ from .util import (
     s3enc,
     s3dec,
     statdir,
+    s2hms,
 )
 from .mtag import MTag
 from .authsrv import AuthSrv
@@ -64,7 +66,7 @@ class Up2k(object):
         self.flags = {}
         self.cur = {}
         self.mtag = None
-        self.n_mtag_tags_added = -1
+        self.pending_tags = None
 
         self.mem_cur = None
         self.sqlite_ver = None
@@ -107,6 +109,10 @@ class Up2k(object):
             thr.daemon = True
             thr.start()
 
+            thr = threading.Thread(target=self._run_all_mtp)
+            thr.daemon = True
+            thr.start()
+
     def log(self, msg, c=0):
         self.log_func("up2k", msg + "\033[K", c)
 
@@ -219,6 +225,16 @@ class Up2k(object):
 
             _, flags = self._expr_idx_filter(flags)
 
+            ft = "\033[0;32m{}{:.0}"
+            ff = "\033[0;35m{}{:.0}"
+            fv = "\033[0;36m{}:\033[1;30m{}"
+            a = [
+                (ft if v is True else ff if v is False else fv).format(k, str(v))
+                for k, v in flags.items()
+            ]
+            if a:
+                self.log(" ".join(sorted(a)) + "\033[0m")
+
             reg = {}
             path = os.path.join(ptop, ".hist", "up2k.snap")
             if "e2d" in flags and os.path.exists(path):
@@ -272,9 +288,12 @@ class Up2k(object):
         dbw = [reg[0], 0, time.time()]
         self.pp.n = next(dbw[0].execute("select count(w) from up"))[0]
 
-        # can be symlink so don't `and d.startswith(top)``
-        excl = set([d.realpath for d in all_vols if d != vol])
-        n_add = self._build_dir(dbw, top, excl, top)
+        excl = [
+            vol.realpath + "/" + d.vpath[len(vol.vpath) :].lstrip("/")
+            for d in all_vols
+            if d != vol and (d.vpath.startswith(vol.vpath + "/") or not vol.vpath)
+        ]
+        n_add = self._build_dir(dbw, top, set(excl), top)
         n_rm = self._drop_lost(dbw[0], top)
         if dbw[1]:
             self.log("commit {} new files".format(dbw[1]))
@@ -435,18 +454,7 @@ class Up2k(object):
 
             mpool = False
             if self.mtag.prefer_mt and not self.args.no_mtag_mt:
-                # mp.pool.ThreadPool and concurrent.futures.ThreadPoolExecutor
-                # both do crazy runahead so lets reinvent another wheel
-                nw = os.cpu_count() if hasattr(os, "cpu_count") else 4
-                if self.n_mtag_tags_added == -1:
-                    self.log("using {}x {}".format(nw, self.mtag.backend))
-                    self.n_mtag_tags_added = 0
-
-                mpool = Queue(nw)
-                for _ in range(nw):
-                    thr = threading.Thread(target=self._tag_thr, args=(mpool,))
-                    thr.daemon = True
-                    thr.start()
+                mpool = self._start_mpool()
 
             c2 = cur.connection.cursor()
             c3 = cur.connection.cursor()
@@ -457,19 +465,21 @@ class Up2k(object):
                 if c2.execute(q, (w[:16],)).fetchone():
                     continue
 
+                if "mtp" in flags:
+                    q = "insert into mt values (?,'t:mtp','a')"
+                    c2.execute(q, (w[:16],))
+
                 if rd.startswith("//") or fn.startswith("//"):
                     rd, fn = s3dec(rd, fn)
 
                 abspath = os.path.join(ptop, rd, fn)
                 self.pp.msg = "c{} {}".format(n_left, abspath)
-                args = c3, entags, w, abspath
+                args = [entags, w, abspath]
                 if not mpool:
-                    n_tags = self._tag_file(*args)
+                    n_tags = self._tag_file(c3, *args)
                 else:
-                    mpool.put(args)
-                    with self.mutex:
-                        n_tags = self.n_mtag_tags_added
-                        self.n_mtag_tags_added = 0
+                    mpool.put(["mtag"] + args)
+                    n_tags = len(self._flush_mpool(c3))
 
                 n_add += n_tags
                 n_buf += n_tags
@@ -481,17 +491,220 @@ class Up2k(object):
                     last_write = time.time()
                     n_buf = 0
 
-            if mpool:
-                for _ in range(mpool.maxsize):
-                    mpool.put(None)
-
-                mpool.join()
+            self._stop_mpool(mpool, c3)
 
             c3.close()
             c2.close()
 
         return n_add, n_rm, True
 
+    def _flush_mpool(self, wcur):
+        with self.mutex:
+            ret = []
+            for x in self.pending_tags:
+                self._tag_file(wcur, *x)
+                ret.append(x[1])
+
+            self.pending_tags = []
+            return ret
+
+    def _run_all_mtp(self):
+        t0 = time.time()
+        self.mtp_force = {}
+        self.mtp_parsers = {}
+        for ptop, flags in self.flags.items():
+            if "mtp" in flags:
+                self._run_one_mtp(ptop)
+
+        td = time.time() - t0
+        msg = "mtp finished in {:.2f} sec ({})"
+        self.log(msg.format(td, s2hms(td, True)))
+
+    def _run_one_mtp(self, ptop):
+        db_path = os.path.join(ptop, ".hist", "up2k.db")
+        sz0 = os.path.getsize(db_path) // 1024
+
+        entags = self.entags[ptop]
+
+        force = {}
+        timeout = {}
+        parsers = {}
+        for parser in self.flags[ptop]["mtp"]:
+            orig = parser
+            tag, parser = parser.split("=", 1)
+            if tag not in entags:
+                continue
+
+            while True:
+                try:
+                    bp = os.path.expanduser(parser)
+                    if os.path.exists(bp):
+                        parsers[tag] = [bp, timeout.get(tag, 30)]
+                        break
+                except:
+                    pass
+
+                try:
+                    arg, parser = parser.split(",", 1)
+                    arg = arg.lower()
+
+                    if arg == "f":
+                        force[tag] = True
+                        continue
+
+                    if arg.startswith("t"):
+                        timeout[tag] = int(arg[1:])
+                        continue
+
+                    raise Exception()
+
+                except:
+                    self.log("invalid argument: " + orig, 1)
+                    return
+
+        self.mtp_force[ptop] = force
+        self.mtp_parsers[ptop] = parsers
+
+        q = "select count(w) from mt where k = 't:mtp'"
+        with self.mutex:
+            cur = self.cur[ptop]
+            cur = cur.connection.cursor()
+            wcur = cur.connection.cursor()
+            n_left = cur.execute(q).fetchone()[0]
+
+        mpool = self._start_mpool()
+        batch_sz = mpool.maxsize * 3
+        t_prev = time.time()
+        n_prev = n_left
+        n_done = 0
+        to_delete = {}
+        in_progress = {}
+        while True:
+            with self.mutex:
+                q = "select w from mt where k = 't:mtp' limit ?"
+                warks = cur.execute(q, (batch_sz,)).fetchall()
+                warks = [x[0] for x in warks]
+                jobs = []
+                for w in warks:
+                    q = "select rd, fn from up where substr(w,1,16)=? limit 1"
+                    rd, fn = cur.execute(q, (w,)).fetchone()
+                    rd, fn = s3dec(rd, fn)
+                    abspath = os.path.join(ptop, rd, fn)
+
+                    q = "select k from mt where w = ?"
+                    have = cur.execute(q, (w,)).fetchall()
+                    have = [x[0] for x in have]
+
+                    if ".dur" not in have and ".dur" in entags:
+                        # skip non-audio
+                        to_delete[w] = True
+                        n_left -= 1
+                        continue
+
+                    if w in in_progress:
+                        continue
+
+                    task_parsers = {
+                        k: v for k, v in parsers.items() if k in force or k not in have
+                    }
+                    jobs.append([task_parsers, None, w, abspath])
+                    in_progress[w] = True
+
+            done = self._flush_mpool(wcur)
+
+            with self.mutex:
+                for w in done:
+                    to_delete[w] = True
+                    in_progress.pop(w)
+                    n_done += 1
+
+                for w in to_delete.keys():
+                    q = "delete from mt where w = ? and k = 't:mtp'"
+                    cur.execute(q, (w,))
+
+                to_delete = {}
+
+            if not warks:
+                break
+
+            if not jobs:
+                continue
+
+            try:
+                now = time.time()
+                s = ((now - t_prev) / (n_prev - n_left)) * n_left
+                h, s = divmod(s, 3600)
+                m, s = divmod(s, 60)
+                n_prev = n_left
+                t_prev = now
+            except:
+                h = 1
+                m = 1
+
+            msg = "mtp: {} done, {} left, eta {}h {:02d}m"
+            with self.mutex:
+                msg = msg.format(n_done, n_left, int(h), int(m))
+                self.log(msg, c=6)
+
+            for j in jobs:
+                n_left -= 1
+                mpool.put(j)
+
+            with self.mutex:
+                cur.connection.commit()
+
+        done = self._stop_mpool(mpool, wcur)
+        with self.mutex:
+            for w in done:
+                q = "delete from mt where w = ? and k = 't:mtp'"
+                cur.execute(q, (w,))
+
+            cur.connection.commit()
+            if n_done:
+                self.vac(cur, db_path, n_done, 0, sz0)
+
+            wcur.close()
+            cur.close()
+
+    def _start_mpool(self):
+        if WINDOWS and False:
+            nah = open(os.devnull, "wb")
+            wmic = "processid={}".format(os.getpid())
+            wmic = ["wmic", "process", "where", wmic, "call", "setpriority"]
+            sp.call(wmic + ["below normal"], stdout=nah, stderr=nah)
+
+        # mp.pool.ThreadPool and concurrent.futures.ThreadPoolExecutor
+        # both do crazy runahead so lets reinvent another wheel
+        nw = os.cpu_count() if hasattr(os, "cpu_count") else 4
+        if self.pending_tags is None:
+            self.log("using {}x {}".format(nw, self.mtag.backend))
+            self.pending_tags = []
+
+        mpool = Queue(nw)
+        for _ in range(nw):
+            thr = threading.Thread(target=self._tag_thr, args=(mpool,))
+            thr.daemon = True
+            thr.start()
+
+        return mpool
+
+    def _stop_mpool(self, mpool, wcur):
+        if not mpool:
+            return
+
+        for _ in range(mpool.maxsize):
+            mpool.put(None)
+
+        mpool.join()
+        done = self._flush_mpool(wcur)
+        if WINDOWS and False:
+            nah = open(os.devnull, "wb")
+            wmic = "processid={}".format(os.getpid())
+            wmic = ["wmic", "process", "where", wmic, "call", "setpriority"]
+            sp.call(wmic + ["below normal"], stdout=nah, stderr=nah)
+
+        return done
+
     def _tag_thr(self, q):
         while True:
             task = q.get()
@@ -500,24 +713,47 @@ class Up2k(object):
                 return
 
             try:
-                write_cur, entags, wark, abspath = task
-                tags = self.mtag.get(abspath)
+                parser, entags, wark, abspath = task
+                if parser == "mtag":
+                    tags = self.mtag.get(abspath)
+                else:
+                    tags = self.mtag.get_bin(parser, abspath)
+                    vtags = [
+                        "\033[36m{} \033[33m{}".format(k, v) for k, v in tags.items()
+                    ]
+                    self.log("{}\033[0m [{}]".format(" ".join(vtags), abspath))
+
                 with self.mutex:
-                    n = self._tag_file(write_cur, entags, wark, abspath, tags)
-                    self.n_mtag_tags_added += n
+                    self.pending_tags.append([entags, wark, abspath, tags])
             except:
                 ex = traceback.format_exc()
+                if parser == "mtag":
+                    parser = self.mtag.backend
+
                 msg = "{} failed to read tags from {}:\n{}"
-                self.log(msg.format(self.mtag.backend, abspath, ex), c=3)
+                self.log(msg.format(parser, abspath, ex), c=3)
 
             q.task_done()
 
     def _tag_file(self, write_cur, entags, wark, abspath, tags=None):
-        tags = tags or self.mtag.get(abspath)
-        tags = {k: v for k, v in tags.items() if k in entags}
+        if tags is None:
+            tags = self.mtag.get(abspath)
+
+        if entags:
+            tags = {k: v for k, v in tags.items() if k in entags}
+            if not tags:
+                # indicate scanned without tags
+                tags = {"x": 0}
+
         if not tags:
-            # indicate scanned without tags
-            tags = {"x": 0}
+            return 0
+
+        for k in tags.keys():
+            q = "delete from mt where w = ? and ({})".format(
+                " or ".join(["k = ?"] * len(tags))
+            )
+            args = [wark[:16]] + list(tags.keys())
+            write_cur.execute(q, tuple(args))
 
         ret = 0
         for k, v in tags.items():
@@ -529,6 +765,7 @@ class Up2k(object):
 
     def _orz(self, db_path):
         return sqlite3.connect(db_path, check_same_thread=False).cursor()
+        # x.set_trace_callback(trace)
 
     def _open_db(self, db_path):
         existed = os.path.exists(db_path)
@@ -1053,8 +1290,20 @@ class Up2k(object):
     def _tagger(self):
         while True:
             ptop, wark, rd, fn = self.tagq.get()
+            if "e2t" not in self.flags[ptop]:
+                continue
+
             abspath = os.path.join(ptop, rd, fn)
-            self.log("tagging " + abspath)
+            tags = self.mtag.get(abspath)
+            ntags1 = len(tags)
+            if self.mtp_parsers.get(ptop, {}):
+                parser = {
+                    k: v
+                    for k, v in self.mtp_parsers[ptop].items()
+                    if k in self.mtp_force[ptop] or k not in tags
+                }
+                tags.update(self.mtag.get_bin(parser, abspath))
+
             with self.mutex:
                 cur = self.cur[ptop]
                 if not cur:
@@ -1066,11 +1315,11 @@ class Up2k(object):
                     self.log("no entags okay.jpg", c=3)
                     continue
 
-                if "e2t" in self.flags[ptop]:
-                    self._tag_file(cur, entags, wark, abspath)
-
+                self._tag_file(cur, entags, wark, abspath, tags)
                 cur.connection.commit()
 
+            self.log("tagged {} ({}+{})".format(abspath, ntags1, len(tags) - ntags1))
+
     def _hasher(self):
         while True:
             ptop, rd, fn = self.hashq.get()

copyparty/util.py

@@ -10,6 +10,7 @@ import select
 import struct
 import hashlib
 import platform
+import traceback
 import threading
 import mimetypes
 import contextlib
@@ -56,11 +57,58 @@ HTTPCODE = {
     413: "Payload Too Large",
     416: "Requested Range Not Satisfiable",
     422: "Unprocessable Entity",
+    429: "Too Many Requests",
     500: "Internal Server Error",
     501: "Not Implemented",
 }
 
 
+IMPLICATIONS = [
+    ["e2dsa", "e2ds"],
+    ["e2ds", "e2d"],
+    ["e2tsr", "e2ts"],
+    ["e2ts", "e2t"],
+    ["e2t", "e2d"],
+]
+
+
+REKOBO_KEY = {
+    v: ln.split(" ", 1)[0]
+    for ln in """
+1B 6d B
+2B 7d Gb F#
+3B 8d Db C#
+4B 9d Ab G#
+5B 10d Eb D#
+6B 11d Bb A#
+7B 12d F
+8B 1d C
+9B 2d G
+10B 3d D
+11B 4d A
+12B 5d E
+1A 6m Abm G#m
+2A 7m Ebm D#m
+3A 8m Bbm A#m
+4A 9m Fm
+5A 10m Cm
+6A 11m Gm
+7A 12m Dm
+8A 1m Am
+9A 2m Em
+10A 3m Bm
+11A 4m Gbm F#m
+12A 5m Dbm C#m
+""".strip().split(
+        "\n"
+    )
+    for v in ln.strip().split(" ")[1:]
+    if v
+}
+
+REKOBO_LKEY = {k.lower(): v for k, v in REKOBO_KEY.items()}
+
+
 class Counter(object):
     def __init__(self, v=0):
         self.v = v
@@ -139,6 +187,31 @@ def nuprint(msg):
     uprint("{}\n".format(msg))
 
 
+def rice_tid():
+    tid = threading.current_thread().ident
+    c = struct.unpack(b"B" * 5, struct.pack(b">Q", tid)[-5:])
+    return "".join("\033[1;37;48;5;{}m{:02x}".format(x, x) for x in c) + "\033[0m"
+
+
+def trace(*args, **kwargs):
+    t = time.time()
+    stack = "".join(
+        "\033[36m{}\033[33m{}".format(x[0].split(os.sep)[-1][:-3], x[1])
+        for x in traceback.extract_stack()[3:-1]
+    )
+    parts = ["{:.6f}".format(t), rice_tid(), stack]
+
+    if args:
+        parts.append(repr(args))
+
+    if kwargs:
+        parts.append(repr(kwargs))
+
+    msg = "\033[0m ".join(parts)
+    # _tracebuf.append(msg)
+    nuprint(msg)
+
+
 @contextlib.contextmanager
 def ren_open(fname, *args, **kwargs):
     fdir = kwargs.pop("fdir", None)
@@ -477,6 +550,16 @@ def get_spd(nbyte, t0, t=None):
     return "{} \033[0m{}/s\033[0m".format(s1, s2)
 
 
+def s2hms(s, optional_h=False):
+    s = int(s)
+    h, s = divmod(s, 3600)
+    m, s = divmod(s, 60)
+    if not h and optional_h:
+        return "{}:{:02}".format(m, s)
+
+    return "{}:{:02}:{:02}".format(h, m, s)
+
+
 def undot(path):
     ret = []
     for node in path.split("/"):
@@ -493,11 +576,12 @@ def undot(path):
     return "/".join(ret)
 
 
-def sanitize_fn(fn):
-    fn = fn.replace("\\", "/").split("/")[-1]
+def sanitize_fn(fn, ok=""):
+    if "/" not in ok:
+        fn = fn.replace("\\", "/").split("/")[-1]
 
     if WINDOWS:
-        for bad, good in [
+        for bad, good in [x for x in [
             ["<", "＜"],
             [">", "＞"],
             [":", "："],
@@ -507,7 +591,7 @@ def sanitize_fn(fn):
             ["|", "｜"],
             ["?", "？"],
             ["*", "＊"],
-        ]:
+        ] if x[0] not in ok]:
             fn = fn.replace(bad, good)
 
         bad = ["con", "prn", "aux", "nul"]
@@ -697,6 +781,16 @@ def read_socket_chunked(sr, log=None):
         sr.recv(2)  # \r\n after each chunk too
 
 
+def yieldfile(fn):
+    with open(fsenc(fn), "rb", 512 * 1024) as f:
+        while True:
+            buf = f.read(64 * 1024)
+            if not buf:
+                break
+
+            yield buf
+
+
 def hashcopy(actor, fin, fout):
     u32_lim = int((2 ** 31) * 0.9)
     hashobj = hashlib.sha512()

copyparty/web/browser.css

@@ -182,6 +182,11 @@ a, #files tbody div a:last-child {
 	color: #840;
 	text-shadow: 0 0 .3em #b80;
 }
+#files tbody tr.sel td {
+	background: #80b;
+	color: #fff;
+	border-color: #a3d;
+}
 #blocked {
 	position: fixed;
 	top: 0;
@@ -268,6 +273,25 @@ a, #files tbody div a:last-child {
 	padding: .2em 0 0 .07em;
 	color: #fff;
 }
+#wtoggle>span {
+	display: none;
+}
+#wtoggle.sel {
+	width: 4.27em;
+}
+#wtoggle.sel>span {
+	display: inline-block;
+	line-height: 0;
+}
+#wtoggle.sel>span a {
+	font-size: .4em;
+	margin: -.3em 0;
+	position: relative;
+	display: inline-block;
+}
+#wtoggle.sel>span #selzip {
+	top: -.6em;
+}
 #barpos,
 #barbuf {
 	position: absolute;
@@ -405,7 +429,7 @@ input[type="checkbox"]:checked+label {
 
 
 
-#op_search table {
+#srch_form {
 	border: 1px solid #3a3a3a;
 	box-shadow: 0 0 1em #222 inset;
 	background: #2d2d2d;
@@ -414,14 +438,25 @@ input[type="checkbox"]:checked+label {
 	margin-bottom: 0;
 	padding: 0 .5em .5em 0;
 }
+#srch_form table {
+	display: inline-block;
+}
 #srch_form td {
 	padding: .6em .6em;
 }
+#srch_form td:first-child {
+	width: 3em;
+	padding-right: .2em;
+	text-align: right;
+}
 #op_search input {
 	margin: 0;
 }
 #srch_q {
 	white-space: pre;
+	color: #f80;
+	height: 1em;
+	margin: .2em 0 -1em 1.6em;
 }
 #files td div span {
 	color: #fff;
@@ -482,6 +517,7 @@ input[type="checkbox"]:checked+label {
 	left: -1.7em;
 	width: calc(100% + 1.3em);
 }
+.tglbtn,
 #tree>a+a {
 	padding: .2em .4em;
 	font-size: 1.2em;
@@ -492,9 +528,11 @@ input[type="checkbox"]:checked+label {
 	position: relative;
 	top: -.2em;
 }
+.tglbtn:hover,
 #tree>a+a:hover {
 	background: #805;
 }
+.tglbtn.on,
 #tree>a+a.on {
 	background: #fc4;
 	color: #400;
@@ -601,7 +639,10 @@ input[type="checkbox"]:checked+label {
 	max-width: none;
 	margin-right: 1.5em;
 }
-#key_notation>span {
+#op_cfg>div>a {
+	line-height: 2em;
+}
+#op_cfg>div>span {
 	display: inline-block;
 	padding: .2em .4em;
 }
@@ -624,6 +665,9 @@ input[type="checkbox"]:checked+label {
 	top: 6em;
 	right: 1.5em;
 }
+#ops:hover #opdesc.off {
+	display: none;
+}
 #opdesc code {
 	background: #3c3c3c;
 	padding: .2em .3em;
@@ -631,4 +675,4 @@ input[type="checkbox"]:checked+label {
 	border-radius: .3em;
 	font-family: monospace, monospace;
 	line-height: 2em;
-}
+}

copyparty/web/browser.html

@@ -29,9 +29,9 @@
 
     <div id="op_search" class="opview">
         {%- if have_tags_idx %}
-        <table id="srch_form" class="tags"></table>
+        <div id="srch_form" class="tags"></div>
         {%- else %}
-        <table id="srch_form"></table>
+        <div id="srch_form"></div>
         {%- endif %}
         <div id="srch_q"></div>
     </div>
@@ -41,6 +41,12 @@
     <div id="op_cfg" class="opview opbox">
         <h3>key notation</h3>
         <div id="key_notation"></div>
+        {%- if have_zip %}
+        <h3>folder download</h3>
+        <div id="arc_fmt"></div>
+        {%- endif %}
+        <h3>tooltips</h3>
+        <div><a id="tooltips" class="tglbtn" href="#">enable</a></div>
     </div>
     
     <h1 id="path">
@@ -54,7 +60,7 @@
         <a href="#" id="detree">🍞...</a>
         <a href="#" step="2" id="twobytwo">+</a>
         <a href="#" step="-2" id="twig">&ndash;</a>
-        <a href="#" id="dyntree">a</a>
+        <a href="#" class="tglbtn" id="dyntree">a</a>
         <ul id="treeul"></ul>
         <div id="thx_ff">&nbsp;</div>
     </div>
@@ -66,18 +72,18 @@
     <table id="files">
         <thead>
             <tr>
-                <th></th>
-                <th><span>File Name</span></th>
-                <th sort="int"><span>Size</span></th>
+                <th name="lead"><span>c</span></th>
+                <th name="href"><span>File Name</span></th>
+                <th name="sz" sort="int"><span>Size</span></th>
                 {%- for k in taglist %}
                     {%- if k.startswith('.') %}
-                        <th sort="int"><span>{{ k[1:] }}</span></th>
+                        <th name="tags/{{ k }}" sort="int"><span>{{ k[1:] }}</span></th>
                     {%- else %}
-                        <th><span>{{ k[0]|upper }}{{ k[1:] }}</span></th>
+                        <th name="tags/{{ k }}"><span>{{ k[0]|upper }}{{ k[1:] }}</span></th>
                     {%- endif %}
                 {%- endfor %}
-                <th><span>T</span></th>
-                <th><span>Date</span></th>
+                <th name="ext"><span>T</span></th>
+                <th name="ts"><span>Date</span></th>
             </tr>
         </thead>
         <tbody>
@@ -106,7 +112,14 @@
     {%- endif %}
 
     <div id="widget">
-        <div id="wtoggle">♫</div>
+        <div id="wtoggle">
+            <span>
+                <a href="#" id="selall">sel.<br />all</a>
+                <a href="#" id="selinv">sel.<br />inv.</a>
+                <a href="#" id="selzip">zip</a>
+            </span>
+            ♫
+        </div>
         <div id="widgeti">
             <div id="pctl"><a href="#" id="bprev">⏮</a><a href="#" id="bplay">▶</a><a href="#" id="bnext">⏭</a></div>
             <canvas id="pvol" width="288" height="38"></canvas>

copyparty/web/browser.js

@@ -6,22 +6,19 @@ function dbg(msg) {
 	ebi('path').innerHTML = msg;
 }
 
-makeSortable(ebi('files'), reload_mp);
-
 
 // extract songs + add play column
-function init_mp() {
-	var tracks = [];
-	var ret = {
-		'au': null,
-		'au_native': null,
-		'au_ogvjs': null,
-		'tracks': tracks,
-		'cover_url': ''
-	};
-	var re_audio = /\.(opus|ogg|m4a|aac|mp3|wav|flac)$/i;
+function MPlayer() {
+	this.id = new Date().getTime();
+	this.au = null;
+	this.au_native = null;
+	this.au_ogvjs = null;
+	this.cover_url = '';
+	this.tracks = {};
+	this.order = [];
 
-	var trs = ebi('files').getElementsByTagName('tbody')[0].getElementsByTagName('tr');
+	var re_audio = /\.(opus|ogg|m4a|aac|mp3|wav|flac)$/i;
+	var trs = document.querySelectorAll('#files tbody tr');
 	for (var a = 0, aa = trs.length; a < aa; a++) {
 		var tds = trs[a].getElementsByTagName('td');
 		var link = tds[1].getElementsByTagName('a');
@@ -30,37 +27,48 @@ function init_mp() {
 
 		var m = re_audio.exec(url);
 		if (m) {
-			var ntrack = tracks.length;
-			tracks.push(url);
-
-			tds[0].innerHTML = '<a id="trk' + ntrack + '" href="#trk' + ntrack + '" class="play">play</a></td>';
+			var tid = link.getAttribute('id');
+			this.order.push(tid);
+			this.tracks[tid] = url;
+			tds[0].innerHTML = '<a id="a' + tid + '" href="#a' + tid + '" class="play">play</a></td>';
+			ebi('a' + tid).onclick = ev_play;
 		}
 	}
 
-	for (var a = 0, aa = tracks.length; a < aa; a++)
-		ebi('trk' + a).onclick = ev_play;
-
-	ret.vol = sread('vol');
-	if (ret.vol !== null)
-		ret.vol = parseFloat(ret.vol);
+	this.vol = sread('vol');
+	if (this.vol !== null)
+		this.vol = parseFloat(this.vol);
 	else
-		ret.vol = 0.5;
+		this.vol = 0.5;
 
-	ret.expvol = function () {
-		return 0.5 * ret.vol + 0.5 * ret.vol * ret.vol;
+	this.expvol = function () {
+		return 0.5 * this.vol + 0.5 * this.vol * this.vol;
 	};
 
-	ret.setvol = function (vol) {
-		ret.vol = Math.max(Math.min(vol, 1), 0);
+	this.setvol = function (vol) {
+		this.vol = Math.max(Math.min(vol, 1), 0);
 		swrite('vol', vol);
 
-		if (ret.au)
-			ret.au.volume = ret.expvol();
+		if (this.au)
+			this.au.volume = this.expvol();
 	};
 
-	return ret;
+	this.read_order = function () {
+		var order = [];
+		var links = document.querySelectorAll('#files>tbody>tr>td:nth-child(1)>a');
+		for (var a = 0, aa = links.length; a < aa; a++) {
+			var tid = links[a].getAttribute('id');
+			if (!tid || tid.indexOf('af-') !== 0)
+				continue;
+
+			order.push(tid.slice(1));
+		}
+		this.order = order;
+	};
 }
-var mp = init_mp();
+addcrc();
+var mp = new MPlayer();
+makeSortable(ebi('files'), mp.read_order.bind(mp));
 
 
 // toggle player widget
@@ -180,10 +188,31 @@ var pbar = (function () {
 		pctx.setTransform(scale, 0, 0, scale, 0, 0);
 		pctx.clearRect(0, 0, sw, sh);
 
+		pctx.fillStyle = 'rgba(204,255,128,0.15)';
+		for (var p = 1, mins = mp.au.duration / 10; p <= mins; p++)
+			pctx.fillRect(Math.floor(sm * p * 10), 0, 2, sh);
+
+		pctx.fillStyle = '#9b7';
+		pctx.fillStyle = 'rgba(192,255,96,0.5)';
+		for (var p = 1, mins = mp.au.duration / 60; p <= mins; p++)
+			pctx.fillRect(Math.floor(sm * p * 60), 0, 2, sh);
+
 		var w = 8;
 		var x = sm * mp.au.currentTime;
 		pctx.fillStyle = '#573'; pctx.fillRect((x - w / 2) - 1, 0, w + 2, sh);
 		pctx.fillStyle = '#dfc'; pctx.fillRect((x - w / 2), 0, 8, sh);
+
+		pctx.fillStyle = '#fff';
+		pctx.font = '1em sans-serif';
+		var txt = s2ms(mp.au.duration);
+		var tw = pctx.measureText(txt).width;
+		pctx.fillText(txt, sw - (tw + 8), sh / 3 * 2);
+
+		txt = s2ms(mp.au.currentTime);
+		tw = pctx.measureText(txt).width;
+		var gw = pctx.measureText("88:88::").width;
+		var xt = x < sw / 2 ? (x + 8) : (Math.min(sw - gw, x - 8) - tw);
+		pctx.fillText(txt, xt, sh / 3 * 2);
 	};
 	return r;
 })();
@@ -277,18 +306,41 @@ var vbar = (function () {
 })();
 
 
+function seek_au_mul(mul) {
+	if (mp.au)
+		seek_au_sec(mp.au.duration * mul);
+}
+
+function seek_au_sec(seek) {
+	if (!mp.au)
+		return;
+
+	console.log('seek: ' + seek);
+	if (!isFinite(seek))
+		return;
+
+	mp.au.currentTime = seek;
+
+	if (mp.au === mp.au_native)
+		// hack: ogv.js breaks on .play() during playback
+		mp.au.play();
+};
+
+
+function song_skip(n) {
+	var tid = null;
+	if (mp.au)
+		tid = mp.au.tid;
+
+	if (tid !== null)
+		play(mp.order.indexOf(tid) + n);
+	else
+		play(mp.order[0]);
+};
+
+
 // hook up the widget buttons
 (function () {
-	var bskip = function (n) {
-		var tid = null;
-		if (mp.au)
-			tid = mp.au.tid;
-
-		if (tid !== null)
-			play(tid + n);
-		else
-			play(0);
-	};
 	ebi('bplay').onclick = function (e) {
 		ev(e);
 		if (mp.au) {
@@ -302,31 +354,20 @@ var vbar = (function () {
 	};
 	ebi('bprev').onclick = function (e) {
 		ev(e);
-		bskip(-1);
+		song_skip(-1);
 	};
 	ebi('bnext').onclick = function (e) {
 		ev(e);
-		bskip(1);
+		song_skip(1);
 	};
 	ebi('barpos').onclick = function (e) {
 		if (!mp.au) {
-			//dbg((new Date()).getTime());
 			return play(0);
 		}
 
 		var rect = pbar.pcan.getBoundingClientRect();
 		var x = e.clientX - rect.left;
-		var mul = x * 1.0 / rect.width;
-		var seek = mp.au.duration * mul;
-		console.log('seek: ' + seek);
-		if (!isFinite(seek))
-			return;
-
-		mp.au.currentTime = seek;
-
-		if (mp.au === mp.au_native)
-			// hack: ogv.js breaks on .play() during playback
-			mp.au.play();
+		seek_au_mul(x * 1.0 / rect.width);
 	};
 })();
 
@@ -359,7 +400,7 @@ var vbar = (function () {
 				var len = mp.au.duration;
 				if (pos > 0 && pos > len - 0.1) {
 					last_skip_url = mp.au.src;
-					play(mp.au.tid + 1);
+					song_skip(1);
 				}
 			}
 		}
@@ -372,7 +413,7 @@ var vbar = (function () {
 // event from play button next to a file in the list
 function ev_play(e) {
 	ev(e);
-	play(parseInt(this.getAttribute('id').substr(3)));
+	play(this.getAttribute('id').slice(1));
 	return false;
 }
 
@@ -394,18 +435,24 @@ catch (ex) { }
 
 // plays the tid'th audio file on the page
 function play(tid, call_depth) {
-	if (mp.tracks.length == 0)
+	if (mp.order.length == 0)
 		return alert('no audio found wait what');
 
-	while (tid >= mp.tracks.length)
-		tid -= mp.tracks.length;
+	var tn = tid;
+	if ((tn + '').indexOf('f-') === 0)
+		tn = mp.order.indexOf(tn);
 
-	while (tid < 0)
-		tid += mp.tracks.length;
+	while (tn >= mp.order.length)
+		tn -= mp.order.length;
+
+	while (tn < 0)
+		tn += mp.order.length;
+
+	tid = mp.order[tn];
 
 	if (mp.au) {
 		mp.au.pause();
-		setclass('trk' + mp.au.tid, 'play');
+		setclass('a' + mp.au.tid, 'play');
 	}
 
 	// ogv.js breaks on .play() unless directly user-triggered
@@ -449,7 +496,7 @@ function play(tid, call_depth) {
 	mp.au.tid = tid;
 	mp.au.src = url;
 	mp.au.volume = mp.expvol();
-	var oid = 'trk' + tid;
+	var oid = 'a' + tid;
 	setclass(oid, 'play act');
 	var trs = ebi('files').getElementsByTagName('tbody')[0].getElementsByTagName('tr');
 	for (var a = 0, aa = trs.length; a < aa; a++) {
@@ -480,8 +527,8 @@ function play(tid, call_depth) {
 	catch (ex) {
 		alert('playback failed: ' + ex);
 	}
-	setclass('trk' + mp.au.tid, 'play');
-	setTimeout('play(' + (mp.au.tid + 1) + ');', 500);
+	setclass(oid, 'play');
+	setTimeout('song_skip(1));', 500);
 }
 
 
@@ -560,12 +607,73 @@ function autoplay_blocked() {
 // autoplay linked track
 (function () {
 	var v = location.hash;
-	if (v && v.length > 4 && v.indexOf('#trk') === 0)
-		play(parseInt(v.substr(4)));
+	if (v && v.length == 12 && v.indexOf('#af-') === 0)
+		play(v.slice(2));
 })();
 
 
-//widget.open();
+function tree_neigh(n) {
+	var links = document.querySelectorAll('#treeul li>a+a');
+	if (!links.length) {
+		alert('switch to the tree for that');
+		return;
+	}
+	var act = -1;
+	for (var a = 0, aa = links.length; a < aa; a++) {
+		if (links[a].getAttribute('class') == 'hl') {
+			act = a;
+			break;
+		}
+	}
+	a += n;
+	if (a < 0)
+		a = links.length - 1;
+	if (a >= links.length)
+		a = 0;
+
+	links[a].click();
+}
+
+
+function tree_up() {
+	var act = document.querySelector('#treeul a.hl');
+	if (!act) {
+		alert('switch to the tree for that');
+		return;
+	}
+	if (act.previousSibling.textContent == '-')
+		return act.previousSibling.click();
+
+	act.parentNode.parentNode.parentNode.getElementsByTagName('a')[1].click();
+}
+
+
+document.onkeydown = function (e) {
+	if (document.activeElement != document.body && document.activeElement.nodeName.toLowerCase() != 'a')
+		return;
+
+	var k = e.code, pos = -1;
+	if (k.indexOf('Digit') === 0)
+		pos = parseInt(k.slice(-1)) * 0.1;
+
+	if (pos !== -1)
+		return seek_au_mul(pos);
+
+	var n = k == 'KeyJ' ? -1 : k == 'KeyL' ? 1 : 0;
+	if (n !== 0)
+		return song_skip(n);
+
+	n = k == 'KeyU' ? -10 : k == 'KeyO' ? 10 : 0;
+	if (n !== 0)
+		return mp.au ? seek_au_sec(mp.au.currentTime + n) : true;
+
+	n = k == 'KeyI' ? -1 : k == 'KeyK' ? 1 : 0;
+	if (n !== 0)
+		return tree_neigh(n);
+
+	if (k == 'KeyP')
+		return tree_up();
+};
 
 
 // search
@@ -588,15 +696,19 @@ function autoplay_blocked() {
 	];
 	var oldcfg = [];
 
-	if (document.querySelector('#srch_form.tags'))
+	if (document.querySelector('#srch_form.tags')) {
 		sconf.push(["tags",
 			["tags", "tags", "tags contains &nbsp; (^=start, end=$)", "46"]
 		]);
+		sconf.push(["adv.",
+			["adv", "adv", "key>=1A&nbsp; key<=2B&nbsp; .bpm>165", "46"]
+		]);
+	}
 
-	var html = [];
+	var trs = [];
 	var orig_html = null;
 	for (var a = 0; a < sconf.length; a++) {
-		html.push('<tr><td><br />' + sconf[a][0] + '</td>');
+		var html = ['<tr><td><br />' + sconf[a][0] + '</td>'];
 		for (var b = 1; b < 3; b++) {
 			var hn = "srch_" + sconf[a][b][0];
 			var csp = (sconf[a].length == 2) ? 2 : 1;
@@ -609,6 +721,11 @@ function autoplay_blocked() {
 				break;
 		}
 		html.push('</tr>');
+		trs.push(html);
+	}
+	var html = [];
+	for (var a = 0; a < trs.length; a += 2) {
+		html.push('<table>' + (trs[a].concat(trs[a + 1])).join('\n') + '</table>');
 	}
 	ebi('srch_form').innerHTML = html.join('\n');
 
@@ -617,7 +734,14 @@ function autoplay_blocked() {
 		o[a].oninput = ev_search_input;
 	}
 
+	function srch_msg(err, txt) {
+		var o = ebi('srch_q');
+		o.textContent = txt;
+		o.style.color = err ? '#f09' : '#c90';
+	}
+
 	var search_timeout;
+	var search_in_progress = 0;
 
 	function ev_search_input() {
 		var v = this.value;
@@ -627,10 +751,14 @@ function autoplay_blocked() {
 			chk.checked = ((v + '').length > 0);
 		}
 		clearTimeout(search_timeout);
-		search_timeout = setTimeout(do_search, 100);
+		var now = new Date().getTime();
+		if (now - search_in_progress > 30 * 1000)
+			search_timeout = setTimeout(do_search, 100);
 	}
 
 	function do_search() {
+		search_in_progress = new Date().getTime();
+		srch_msg(false, "searching...");
 		clearTimeout(search_timeout);
 		var params = {};
 		var o = document.querySelectorAll('#op_search input[type="text"]');
@@ -654,9 +782,16 @@ function autoplay_blocked() {
 			return;
 
 		if (this.status !== 200) {
-			alert("http " + this.status + ": " + this.responseText);
+			var msg = this.responseText;
+			if (msg.indexOf('<pre>') === 0)
+				msg = msg.slice(5);
+
+			srch_msg(true, "http " + this.status + ": " + msg);
+			search_in_progress = 0;
 			return;
 		}
+		search_in_progress = 0;
+		srch_msg(false, '');
 
 		var res = JSON.parse(this.responseText),
 			tagord = res.tag_order;
@@ -674,6 +809,7 @@ function autoplay_blocked() {
 			ebi('path').style.display = 'none';
 			ebi('tree').style.display = 'none';
 			ebi('wrap').style.marginLeft = '0';
+			treectl.hidden = true;
 		}
 
 		var html = mk_files_header(tagord);
@@ -696,7 +832,7 @@ function autoplay_blocked() {
 				var k = tagord[b],
 					v = r.tags[k] || "";
 
-				if (k == "dur") {
+				if (k == ".dur") {
 					var sv = s2ms(v);
 					nodes[nodes.length - 1] += '</td><td sortv="' + v + '">' + sv;
 					continue;
@@ -716,6 +852,7 @@ function autoplay_blocked() {
 		ofiles.innerHTML = html.join('\n');
 		ofiles.setAttribute("ts", this.ts);
 		filecols.set_style();
+		mukey.render();
 		reload_browser();
 
 		ebi('unsearch').onclick = unsearch;
@@ -726,6 +863,7 @@ function autoplay_blocked() {
 		ebi('path').style.display = oldcfg[0];
 		ebi('tree').style.display = oldcfg[1];
 		ebi('wrap').style.marginLeft = oldcfg[2];
+		treectl.hidden = false;
 		oldcfg = [];
 		ebi('files').innerHTML = orig_html;
 		orig_html = null;
@@ -735,6 +873,9 @@ function autoplay_blocked() {
 
 
 var treectl = (function () {
+	var treectl = {
+		"hidden": false
+	};
 	var dyn = bcfg_get('dyntree', true);
 	var treesz = icfg_get('treesz', 16);
 	treesz = Math.min(Math.max(treesz, 4), 50);
@@ -768,7 +909,7 @@ var treectl = (function () {
 	}
 
 	function onscroll() {
-		if (!entreed)
+		if (!entreed || treectl.hidden)
 			return;
 
 		var top = ebi('wrap').getBoundingClientRect().top;
@@ -782,7 +923,7 @@ var treectl = (function () {
 	periodic();
 
 	function onresize(e) {
-		if (!entreed)
+		if (!entreed || treectl.hidden)
 			return;
 
 		var q = '#tree';
@@ -944,7 +1085,36 @@ var treectl = (function () {
 		}
 
 		ebi('srv_info').innerHTML = '<span>' + res.srvinf + '</span>';
-		var nodes = res.dirs.concat(res.files);
+		var nodes = res.dirs.concat(res.files),
+			sopts = jread('fsort', []);
+
+		try {
+			for (var a = sopts.length - 1; a >= 0; a--) {
+				var name = sopts[a][0], rev = sopts[a][1], typ = sopts[a][2];
+				if (!name)
+					continue;
+
+				if (name.indexOf('tags/') == -1) {
+					nodes.sort(function (v1, v2) {
+						if (!v1[name]) return -1 * rev;
+						if (!v2[name]) return 1 * rev;
+						return rev * (typ == 'int' ? (v1[name] - v2[name]) : (v1[name].localeCompare(v2[name])));
+					});
+				}
+				else {
+					name = name.slice(5);
+					nodes.sort(function (v1, v2) {
+						if (!v1.tags[name]) return -1 * rev;
+						if (!v2.tags[name]) return 1 * rev;
+						return rev * (typ == 'int' ? (v1.tags[name] - v2.tags[name]) : (v1.tags[name].localeCompare(v2.tags[name])));
+					});
+				}
+			}
+		}
+		catch (ex) {
+			console.log("failed to apply sort config: " + ex);
+		}
+
 		var top = this.top;
 		var html = mk_files_header(res.taglist);
 		html.push('<tbody>');
@@ -957,10 +1127,7 @@ var treectl = (function () {
 				var k = res.taglist[b],
 					v = (r.tags || {})[k] || "";
 
-				if (k[0] == '.')
-					k = k.slice(1);
-
-				if (k == "dur") {
+				if (k == ".dur") {
 					var sv = s2ms(v);
 					ln[ln.length - 1] += '</td><td sortv="' + v + '">' + sv;
 					continue;
@@ -985,6 +1152,7 @@ var treectl = (function () {
 
 		filecols.set_style();
 		mukey.render();
+		msel.render();
 		reload_tree();
 		reload_browser();
 	}
@@ -1060,9 +1228,8 @@ var treectl = (function () {
 		hist_replace(get_evpath() + window.location.hash);
 	}
 
-	return {
-		"onscroll": onscroll
-	}
+	treectl.onscroll = onscroll;
+	return treectl;
 })();
 
 
@@ -1118,27 +1285,47 @@ function apply_perms(perms) {
 }
 
 
+function find_file_col(txt) {
+	var tds = ebi('files').tHead.getElementsByTagName('th');
+	var i = -1;
+	var min = false;
+	for (var a = 0; a < tds.length; a++) {
+		var spans = tds[a].getElementsByTagName('span');
+		if (spans.length && spans[0].textContent == txt) {
+			min = tds[a].getAttribute('class').indexOf('min') !== -1;
+			i = a;
+			break;
+		}
+	}
+
+	if (i == -1)
+		return;
+
+	return [i, min];
+}
+
+
 function mk_files_header(taglist) {
 	var html = [
 		'<thead>',
-		'<th></th>',
-		'<th><span>File Name</span></th>',
-		'<th sort="int"><span>Size</span></th>'
+		'<th name="lead"><span>c</span></th>',
+		'<th name="href"><span>File Name</span></th>',
+		'<th name="sz" sort="int"><span>Size</span></th>'
 	];
 	for (var a = 0; a < taglist.length; a++) {
 		var tag = taglist[a];
 		var c1 = tag.slice(0, 1).toUpperCase();
 		tag = c1 + tag.slice(1);
 		if (c1 == '.')
-			tag = '<th sort="int"><span>' + tag.slice(1);
+			tag = '<th name="tags/' + tag + '" sort="int"><span>' + tag.slice(1);
 		else
-			tag = '<th><span>' + tag;
+			tag = '<th name="tags/' + tag + '"><span>' + tag;
 
 		html.push(tag + '</span></th>');
 	}
 	html = html.concat([
-		'<th><span>T</span></th>',
-		'<th><span>Date</span></th>',
+		'<th name="ext"><span>T</span></th>',
+		'<th name="ts"><span>Date</span></th>',
 		'</thead>',
 	]);
 	return html;
@@ -1215,6 +1402,21 @@ var filecols = (function () {
 		set_style();
 	};
 
+	try {
+		var ci = find_file_col('dur'),
+			i = ci[0],
+			min = ci[1],
+			rows = ebi('files').tBodies[0].rows;
+
+		if (!min)
+			for (var a = 0, aa = rows.length; a < aa; a++) {
+				var c = rows[a].cells[i];
+				if (c)
+					var v = c.textContent = s2ms(c.textContent);
+			}
+	}
+	catch (ex) { }
+
 	return {
 		"add_btns": add_btns,
 		"set_style": set_style,
@@ -1266,7 +1468,7 @@ var mukey = (function () {
 		ev(e);
 		var notation = this.getAttribute('value');
 		load_notation(notation);
-		render();
+		try_render();
 	}
 
 	function load_notation(notation) {
@@ -1281,22 +1483,10 @@ var mukey = (function () {
 	}
 
 	function render() {
-		var tds = ebi('files').tHead.getElementsByTagName('th');
-		var i = -1;
-		var min = false;
-		for (var a = 0; a < tds.length; a++) {
-			var spans = tds[a].getElementsByTagName('span');
-			if (spans.length && spans[0].textContent == 'Key') {
-				min = tds[a].getAttribute('class').indexOf('min') !== -1;
-				i = a;
-				break;
-			}
-		}
-
-		if (i == -1)
-			return;
-
-		var rows = ebi('files').tBodies[0].rows;
+		var ci = find_file_col('Key'),
+			i = ci[0],
+			min = ci[1],
+			rows = ebi('files').tBodies[0].rows;
 
 		if (min)
 			for (var a = 0, aa = rows.length; a < aa; a++) {
@@ -1342,15 +1532,169 @@ var mukey = (function () {
 })();
 
 
+function addcrc() {
+	var links = document.querySelectorAll('#files>tbody>tr>td:nth-child(2)>a');
+	for (var a = 0, aa = links.length; a < aa; a++)
+		if (!links[a].getAttribute('id'))
+			links[a].setAttribute('id', 'f-' + crc32(links[a].textContent));
+}
+
+
 (function () {
+	var tt = bcfg_get('tooltips', true);
+
 	function set_tooltip(e) {
 		ev(e);
-		ebi('opdesc').innerHTML = this.getAttribute('data-desc');
+		var o = ebi('opdesc');
+		o.innerHTML = this.getAttribute('data-desc');
+		o.setAttribute('class', tt ? '' : 'off');
 	}
+
 	var btns = document.querySelectorAll('#ops, #ops>a');
 	for (var a = 0; a < btns.length; a++) {
 		btns[a].onmouseenter = set_tooltip;
 	}
+
+	ebi('tooltips').onclick = function (e) {
+		ev(e);
+		tt = !tt;
+		bcfg_set('tooltips', tt);
+	};
+})();
+
+
+var arcfmt = (function () {
+	if (!ebi('arc_fmt'))
+		return { "render": function () { } };
+
+	var html = [],
+		arcfmts = ["tar", "zip", "zip_dos", "zip_crc"],
+		arcv = ["tar", "zip=utf8", "zip", "zip=crc"];
+
+	for (var a = 0; a < arcfmts.length; a++) {
+		var k = arcfmts[a];
+		html.push(
+			'<span><input type="radio" name="arcfmt" value="' + k + '" id="arcfmt_' + k + '">' +
+			'<label for="arcfmt_' + k + '">' + k + '</label></span>');
+	}
+	ebi('arc_fmt').innerHTML = html.join('\n');
+
+	var fmt = sread("arc_fmt") || "zip";
+	ebi('arcfmt_' + fmt).checked = true;
+
+	function render() {
+		var arg = arcv[arcfmts.indexOf(fmt)],
+			tds = document.querySelectorAll('#files tbody td:first-child a');
+
+		for (var a = 0, aa = tds.length; a < aa; a++) {
+			var o = tds[a], txt = o.textContent, href = o.getAttribute('href');
+			if (txt != 'tar' && txt != 'zip')
+				continue;
+
+			var ofs = href.lastIndexOf('?');
+			if (ofs < 0)
+				throw 'missing arg in url';
+
+			o.setAttribute("href", href.slice(0, ofs + 1) + arg);
+			o.textContent = fmt.split('_')[0];
+		}
+		ebi('selzip').textContent = fmt.split('_')[0];
+		ebi('selzip').setAttribute('fmt', arg);
+	}
+
+	function try_render() {
+		try {
+			render();
+		}
+		catch (ex) {
+			console.log("arcfmt failed: " + ex);
+		}
+	}
+
+	function change_fmt(e) {
+		ev(e);
+		fmt = this.getAttribute('value');
+		swrite("arc_fmt", fmt);
+		try_render();
+	}
+
+	var o = document.querySelectorAll('#arc_fmt input');
+	for (var a = 0; a < o.length; a++) {
+		o[a].onchange = change_fmt;
+	}
+
+	return {
+		"render": try_render
+	};
+})();
+
+
+var msel = (function () {
+	function getsel() {
+		var names = [];
+		var links = document.querySelectorAll('#files tbody tr.sel td:nth-child(2) a');
+		for (var a = 0, aa = links.length; a < aa; a++)
+			names.push(links[a].getAttribute('href').split('/').slice(-1));
+
+		return names;
+	}
+	function selui() {
+		var fun = getsel().length ? "add" : "remove";
+		ebi('wtoggle').classList[fun]('sel');
+	}
+	function seltgl(e) {
+		ev(e);
+		var tr = this.parentNode;
+		tr.classList.toggle('sel');
+		selui();
+	}
+	function evsel(e, fun) {
+		ev(e);
+		var trs = document.querySelectorAll('#files tbody tr');
+		for (var a = 0, aa = trs.length; a < aa; a++)
+			trs[a].classList[fun]('sel');
+		selui();
+	}
+	ebi('selall').onclick = function (e) {
+		evsel(e, "add");
+	};
+	ebi('selinv').onclick = function (e) {
+		evsel(e, "toggle");
+	};
+	ebi('selzip').onclick = function (e) {
+		ev(e);
+		var names = getsel();
+		var arg = ebi('selzip').getAttribute('fmt');
+		var txt = names.join('\n');
+		var frm = document.createElement('form');
+		frm.setAttribute('action', '?' + arg);
+		frm.setAttribute('method', 'post');
+		frm.setAttribute('target', '_blank');
+		frm.setAttribute('enctype', 'multipart/form-data');
+		frm.innerHTML = '<input name="act" value="zip" />' +
+			'<textarea name="files" id="ziptxt"></textarea>';
+		frm.style.display = 'none';
+
+		var oldform = document.querySelector('#widgeti>form');
+		if (oldform)
+			oldform.parentNode.removeChild(oldform);
+
+		ebi('widgeti').appendChild(frm);
+		var obj = ebi('ziptxt');
+		obj.value = txt;
+		console.log(txt);
+		frm.submit();
+	};
+	function render() {
+		var tds = document.querySelectorAll('#files tbody td+td+td');
+		for (var a = 0, aa = tds.length; a < aa; a++) {
+			tds[a].onclick = seltgl;
+		}
+		arcfmt.render();
+	}
+	return {
+		"render": render
+	};
 })();
 
 
@@ -1366,13 +1710,12 @@ function reload_mp() {
 		mp.au = null;
 	}
 	widget.close();
-	mp = init_mp();
+	mp = new MPlayer();
 }
 
 
 function reload_browser(not_mp) {
 	filecols.set_style();
-	makeSortable(ebi('files'), reload_mp);
 
 	var parts = get_evpath().split('/');
 	var rm = document.querySelectorAll('#path>a+a+a');
@@ -1384,7 +1727,7 @@ function reload_browser(not_mp) {
 		link += parts[a] + '/';
 		var o = document.createElement('a');
 		o.setAttribute('href', link);
-		o.innerHTML = parts[a];
+		o.textContent = uricom_dec(parts[a])[0];
 		ebi('path').appendChild(o);
 	}
 
@@ -1396,11 +1739,15 @@ function reload_browser(not_mp) {
 		oo[a].textContent = hsz;
 	}
 
-	if (!not_mp)
+	if (!not_mp) {
+		addcrc();
 		reload_mp();
+		makeSortable(ebi('files'), mp.read_order.bind(mp));
+	}
 
 	if (window['up2k'])
 		up2k.set_fsearch();
 }
 reload_browser(true);
 mukey.render();
+msel.render();

copyparty/web/md.html

@@ -147,7 +147,7 @@ var md_opt = {
 
 	</script>
     <script src="/.cpr/util.js"></script>
-	<script src="/.cpr/deps/marked.full.js"></script>
+	<script src="/.cpr/deps/marked.js"></script>
 	<script src="/.cpr/md.js"></script>
 	{%- if edit %}
 	<script src="/.cpr/md2.js"></script>

copyparty/web/up2k.js

@@ -278,18 +278,27 @@ function up2k_init(have_crypto) {
         }
         else files = e.target.files;
 
-        if (files.length == 0)
+        if (!files || files.length == 0)
             return alert('no files selected??');
 
         more_one_file();
         var bad_files = [];
         var good_files = [];
+        var dirs = [];
         for (var a = 0; a < files.length; a++) {
             var fobj = files[a];
             if (is_itemlist) {
                 if (fobj.kind !== 'file')
                     continue;
 
+                try {
+                    var wi = fobj.webkitGetAsEntry();
+                    if (wi.isDirectory) {
+                        dirs.push(wi);
+                        continue;
+                    }
+                }
+                catch (ex) { }
                 fobj = fobj.getAsFile();
             }
             try {
@@ -300,12 +309,69 @@ function up2k_init(have_crypto) {
                 bad_files.push(fobj.name);
                 continue;
             }
-            good_files.push(fobj);
+            good_files.push([fobj, fobj.name]);
+        }
+        if (dirs) {
+            return read_dirs(null, [], dirs, good_files, bad_files);
+        }
+    }
+
+    function read_dirs(rd, pf, dirs, good, bad) {
+        if (!dirs.length) {
+            if (!pf.length)
+                return gotallfiles(good, bad);
+
+            console.log("retry pf, " + pf.length);
+            setTimeout(function () {
+                read_dirs(rd, pf, dirs, good, bad);
+            }, 50);
+            return;
         }
 
+        if (!rd)
+            rd = dirs[0].createReader();
+
+        rd.readEntries(function (ents) {
+            var ngot = 0;
+            ents.forEach(function (dn) {
+                if (dn.isDirectory) {
+                    dirs.push(dn);
+                }
+                else {
+                    var name = dn.fullPath;
+                    if (name.indexOf('/') === 0)
+                        name = name.slice(1);
+
+                    pf.push(name);
+                    dn.file(function (fobj) {
+                        var idx = pf.indexOf(name);
+                        pf.splice(idx, 1);
+                        try {
+                            if (fobj.size > 0) {
+                                good.push([fobj, name]);
+                                return;
+                            }
+                        }
+                        catch (ex) { }
+                        bad.push(name);
+                    });
+                }
+                ngot += 1;
+            });
+            // console.log("ngot: " + ngot);
+            if (!ngot) {
+                dirs.shift();
+                rd = null;
+            }
+            return read_dirs(rd, pf, dirs, good, bad);
+        });
+    }
+
+    function gotallfiles(good_files, bad_files) {
         if (bad_files.length > 0) {
-            var msg = 'These {0} files (of {1} total) were skipped because they are empty:\n'.format(bad_files.length, files.length);
-            for (var a = 0; a < bad_files.length; a++)
+            var ntot = bad_files.length + good_files.length;
+            var msg = 'These {0} files (of {1} total) were skipped because they are empty:\n'.format(bad_files.length, ntot);
+            for (var a = 0, aa = Math.min(20, bad_files.length); a < aa; a++)
                 msg += '-- ' + bad_files[a] + '\n';
 
             if (files.length - bad_files.length <= 1 && /(android)/i.test(navigator.userAgent))
@@ -315,21 +381,21 @@ function up2k_init(have_crypto) {
         }
 
         var msg = ['upload these ' + good_files.length + ' files?'];
-        for (var a = 0; a < good_files.length; a++)
-            msg.push(good_files[a].name);
+        for (var a = 0, aa = Math.min(20, good_files.length); a < aa; a++)
+            msg.push(good_files[a][1]);
 
         if (ask_up && !fsearch && !confirm(msg.join('\n')))
             return;
 
         for (var a = 0; a < good_files.length; a++) {
-            var fobj = good_files[a];
+            var fobj = good_files[a][0];
             var now = new Date().getTime();
             var lmod = fobj.lastModified || now;
             var entry = {
                 "n": parseInt(st.files.length.toString()),
-                "t0": now,  // TODO remove probably
+                "t0": now,
                 "fobj": fobj,
-                "name": fobj.name,
+                "name": good_files[a][1],
                 "size": fobj.size,
                 "lmod": lmod / 1000,
                 "purl": get_evpath(),

copyparty/web/upload.css

@@ -88,7 +88,7 @@
 	width: 30em;
 }
 #u2conf.has_btn {
-	width: 46em;
+	width: 48em;
 }
 #u2conf * {
 	text-align: center;

copyparty/web/upload.html

@@ -73,7 +73,8 @@
             <div id="u2btn_ct">
                 <div id="u2btn">
                     <span id="u2bm"></span><br />
-                    drop files here<br />
+                    drag/drop files<br />
+                    and folders here<br />
                     (or click me)
                 </div>
             </div>

copyparty/web/util.js

@@ -91,7 +91,29 @@ function import_js(url, cb) {
 }
 
 
-function sortTable(table, col) {
+var crctab = (function () {
+    var c, tab = [];
+    for (var n = 0; n < 256; n++) {
+        c = n;
+        for (var k = 0; k < 8; k++) {
+            c = ((c & 1) ? (0xEDB88320 ^ (c >>> 1)) : (c >>> 1));
+        }
+        tab[n] = c;
+    }
+    return tab;
+})();
+
+
+function crc32(str) {
+    var crc = 0 ^ (-1);
+    for (var i = 0; i < str.length; i++) {
+        crc = (crc >>> 8) ^ crctab[(crc ^ str.charCodeAt(i)) & 0xFF];
+    }
+    return ((crc ^ (-1)) >>> 0).toString(16);
+};
+
+
+function sortTable(table, col, cb) {
     var tb = table.tBodies[0],
         th = table.tHead.rows[0].cells,
         tr = Array.prototype.slice.call(tb.rows, 0),
@@ -100,6 +122,27 @@ function sortTable(table, col) {
         th[a].className = th[a].className.replace(/ *sort-?1 */, " ");
     th[col].className += ' sort' + reverse;
     var stype = th[col].getAttribute('sort');
+    try {
+        var nrules = [], rules = jread("fsort", []);
+        rules.unshift([th[col].getAttribute('name'), reverse, stype || '']);
+        for (var a = 0; a < rules.length; a++) {
+            var add = true;
+            for (var b = 0; b < a; b++)
+                if (rules[a][0] == rules[b][0])
+                    add = false;
+
+            if (add)
+                nrules.push(rules[a]);
+
+            if (nrules.length >= 10)
+                break;
+        }
+        jwrite("fsort", nrules);
+    }
+    catch (ex) {
+        console.log("failed to persist sort rules, resetting: " + ex);
+        jwrite("fsort", null);
+    }
     var vl = [];
     for (var a = 0; a < tr.length; a++) {
         var cell = tr[a].cells[col];
@@ -127,6 +170,7 @@ function sortTable(table, col) {
         return reverse * (a.localeCompare(b));
     });
     for (i = 0; i < tr.length; ++i) tb.appendChild(tr[vl[i][1]]);
+    if (cb) cb();
 }
 function makeSortable(table, cb) {
     var th = table.tHead, i;
@@ -136,9 +180,7 @@ function makeSortable(table, cb) {
     while (--i >= 0) (function (i) {
         th[i].onclick = function (e) {
             ev(e);
-            sortTable(table, i);
-            if (cb)
-                cb();
+            sortTable(table, i, cb);
         };
     }(i));
 }
@@ -195,7 +237,10 @@ function goto(dest) {
     goto();
     var op = sread('opmode');
     if (op !== null && op !== '.')
-        goto(op);
+        try {
+            goto(op);
+        }
+        catch (ex) { }
 })();
 
 
@@ -275,6 +320,7 @@ function unix2iso(ts) {
 
 
 function s2ms(s) {
+    s = Math.floor(s);
     var m = Math.floor(s / 60);
     return m + ":" + ("0" + (s - m * 60)).slice(-2);
 }
@@ -361,8 +407,10 @@ function bcfg_upd_ui(name, val) {
 
     if (o.getAttribute('type') == 'checkbox')
         o.checked = val;
-    else if (o)
-        o.setAttribute('class', val ? 'on' : '');
+    else if (o) {
+        var fun = val ? 'add' : 'remove';
+        o.classList[fun]('on');
+    }
 }
 
 

docs/music-analysis.sh

@@ -0,0 +1,242 @@
+#!/bin/bash
+echo please dont actually run this as a scriopt
+exit 1
+
+
+# dependency-heavy, not particularly good fit
+pacman -S llvm10
+python3 -m pip install --user librosa
+git clone https://github.com/librosa/librosa.git
+
+
+# correct bpm for tracks with bad tags
+br='
+/Trip Trip Trip\(Hardcore Edit\).mp3/ {v=176}
+/World!!.BIG_SOS/ {v=175}
+/\/08\..*\(BIG_SOS Bootleg\)\.mp3/ {v=175}
+/もってけ！セーラ服.Asterisk DnB/ {v=175}
+/Rondo\(Asterisk DnB Re.mp3/ {v=175}
+/Ray Nautica 175 Edit/ {v=175;x="thunk"}
+/TOKIMEKI Language.Jauz/ {v=174}
+/YUPPUN Hardcore Remix\).mp3/ {v=174;x="keeps drifting"}
+/(èâAâï.î╧ûδ|バーチャリアル.狐耶)J-Core Remix\).mp3/ {v=172;x="hard"}
+/lucky train..Freezer/ {v=170}
+/Alf zero Bootleg ReMix/ {v=170}
+/Prisoner of Love.Kacky/ {v=170}
+/火炎 .Qota/ {v=170}
+/\(hu-zin Bootleg\)\.mp3/ {v=170}
+/15. STRAIGHT BET\(Milynn Bootleg\)\.mp3/ {v=170}
+/\/13.*\(Milynn Bootleg\)\.mp3/ {v=167;x="way hard"}
+/COLOR PLANET .10SAI . nijikon Remix\)\.mp3/ {v=165}
+/11\. (朝はご飯派|Æ⌐é═é▓ö╤öh)\.mp3/ {v=162}
+/09\. Where.s the core/ {v=160}
+/PLANET\(Koushif Jersey Club Bootleg\)remaster.mp3/ {v=160;x="starts ez turns bs"}
+/kened Soul - Madeon x Angel Beats!.mp3/ {v=160}
+/Dear Moments\(Mother Harlot Bootleg\)\.mp3/ {v=150}
+/POWER.Ringos UKG/ {v=140}
+/ブルー・フィールド\(Ringos UKG Remix\).mp3/ {v=135}
+/プラチナジェット.Ringo Remix..mp3/ {v=131.2}
+/Mirrorball Love \(TKM Bootleg Mix\).mp3/ {v=130}
+/Photon Melodies \(TKM Bootleg Mix\).mp3/ {v=128}
+/Trap of Love \(TKM Bootleg Mix\).mp3/ {v=128}
+/One Step \(TKM Bootleg Mix\)\.mp3/ {v=126}
+/04 (トリカムイ岩|âgâèâJâÇâCèΓ).mp3/ {v=125}
+/Get your Wish \(NAWN REMIX\)\.mp3/ {v=95}
+/Flicker .Nitro Fun/ {v=92}
+/\/14\..*suicat Remix/ {v=85.5;x="tricky"}
+/Yanagi Nagi - Harumodoki \(EO Remix\)\.mp3/ {v=150}
+/Azure - Nicology\.mp3/ {v=128;x="off by 5 how"}
+'
+
+
+# afun host, collects/grades the results
+runfun() { cores=8; touch run; rm -f /dev/shm/mres.*; t00=$(date +%s); tbc() { bc | sed -r 's/(\.[0-9]{2}).*/\1/'; }; for ((core=0; core<$cores; core++)); do sqlite3 /mnt/Users/ed/Music/.hist/up2k.db 'select dur.w, dur.v, bpm.v from mt bpm join mt dur on bpm.w = dur.w where bpm.k = ".bpm" and dur.k = ".dur" order by dur.w' | uniq -w16 | while IFS=\| read w dur bpm; do sqlite3 /mnt/Users/ed/Music/.hist/up2k.db "select rd, fn from up where substr(w,1,16) = '$w'" | sed -r "s/^/$bpm /"; done | grep mir/cr | tr \| / | awk '{v=$1;sub(/[^ ]+ /,"")} '"$br"' {printf "%s %s\n",v,$0}' | while read bpm fn; do [ -e run ] || break; n=$((n+1)); ncore=$((n%cores)); [ $ncore -eq $core ] || continue; t0=$(date +%s.%N); (afun || exit 1; t=$(date +%s.%N); td=$(echo "scale=3; $t - $t0" | tbc); bd=$(echo "scale=3; $bpm / $py" | tbc); printf '%4s sec, %4s orig, %6s py, %4s div, %s\n' $td $bpm $py $bd "$fn") | tee -a /dev/shm/mres.$ncore; rv=${PIPESTATUS[0]}; [ $rv -eq 0 ] || { echo "FAULT($rv): $fn"; }; done & done; wait 2>/dev/null; cat /dev/shm/mres.* | awk 'function prt(c) {printf "\033[3%sm%s\033[0m\n",c,$0} $8!="div,"{next} $5!~/^[0-9\.]+/{next} {meta=$3;det=$5;div=meta/det} div<0.7{det/=2} div>1.3{det*=2} {idet=sprintf("%.0f",det)} {idiff=idet-meta} meta>idet{idiff=meta-idet} idiff==0{n0++;prt(6);next} idiff==1{n1++;prt(3);next} idiff>10{nx++;prt(1);next} {n10++;prt(5)} END {printf "ok: %d   1off: %2s   (%3s)   10off: %2s   (%3s)   fail: %2s\n",n0,n1,n0+n1,n10,n0+n1+n10,nx}'; te=$(date +%s); echo $((te-t00)) sec spent; }
+
+
+# ok:   8   1off: 62   ( 70)   10off: 86   (156)   fail: 25   # 105 sec,  librosa @ 8c archvm on 3700x w10
+# ok:   4   1off: 59   ( 63)   10off: 65   (128)   fail: 53   # using original tags (bad)
+afun() { ffmpeg -hide_banner -v fatal -nostdin -ss $((dur/3)) -y -i /mnt/Users/ed/Music/"$fn" -t 60 /dev/shm/$core.wav || return 1; py="$(/home/ed/src/librosa/examples/beat_tracker.py /dev/shm/$core.wav x 2>&1 | awk 'BEGIN {v=1} /^Estimated tempo: /{v=$3} END {print v}')"; } runfun
+
+
+# ok: 119   1off:  5   (124)   10off:  8   (132)   fail: 49   # 51 sec,  vamp-example-fixedtempo
+# ok: 109   1off:  4   (113)   10off:  9   (122)   fail: 59   # bad-tags
+afun() { ffmpeg -hide_banner -v fatal -nostdin -ss $((dur/3)) -y -i /mnt/Users/ed/Music/"$fn" -ac 1 -ar 22050 -f f32le /dev/shm/$core.pcm || return 1; py="$(python3 -c 'import vamp; import numpy as np; f = open("/dev/shm/'$core'.pcm", "rb"); d = np.fromfile(f, dtype=np.float32); c = vamp.collect(d, 22050, "vamp-example-plugins:fixedtempo", parameters={"maxdflen":40}); print(c["list"][0]["label"].split(" ")[0])')"; }; runfun
+
+
+# ok: 102   1off: 61   (163)   10off: 12   (175)   fail:  6   # 61 sec,  vamp-qm-tempotracker
+# ok:  80   1off: 48   (128)   10off: 11   (139)   fail: 42   # bad-tags
+afun() { ffmpeg -hide_banner -v fatal -nostdin -ss $((dur/3)) -y -i /mnt/Users/ed/Music/"$fn" -ac 1 -ar 22050 -f f32le /dev/shm/$core.pcm || return 1; py="$(python3 -c 'import vamp; import numpy as np; f = open("/dev/shm/'$core'.pcm", "rb"); d = np.fromfile(f, dtype=np.float32); c = vamp.collect(d, 22050, "qm-vamp-plugins:qm-tempotracker", parameters={"inputtempo":150}); v = [float(x["label"].split(" ")[0]) for x in c["list"] if x["label"]]; v = list(sorted(v))[len(v)//4:-len(v)//4]; print(round(sum(v) / len(v), 1))')"; }; runfun
+
+
+# ok: 133   1off: 32   (165)   10off: 12   (177)   fail:  3   # 51 sec,  vamp-beatroot
+# ok: 101   1off: 22   (123)   10off: 16   (139)   fail: 39   # bad-tags
+# note: some tracks fully fail to analyze (unlike the others which always provide a guess)
+afun() { ffmpeg -hide_banner -v fatal -nostdin -ss $((dur/3)) -y -i /mnt/Users/ed/Music/"$fn" -ac 1 -ar 22050 -f f32le /dev/shm/$core.pcm || return 1; py="$(python3 -c 'import vamp; import numpy as np; f = open("/dev/shm/'$core'.pcm", "rb"); d = np.fromfile(f, dtype=np.float32); c = vamp.collect(d, 22050, "beatroot-vamp:beatroot"); cl=c["list"]; print(round(60*((len(cl)-1)/(float(cl[-1]["timestamp"]-cl[1]["timestamp"]))), 2))')"; }; runfun
+
+
+# ok: 124   1off:  9   (133)   10off: 40   (173)   fail:  8   # 231 sec,  essentia/full
+# ok: 109   1off:  8   (117)   10off: 22   (139)   fail: 42   # bad-tags
+afun() { ffmpeg -hide_banner -v fatal -nostdin -ss $((dur/3)) -y -i /mnt/Users/ed/Music/"$fn" -ac 1 -ar 44100 /dev/shm/$core.wav || return 1; py="$(python3 -c 'import essentia; import essentia.standard as es; fe, fef = es.MusicExtractor(lowlevelStats=["mean", "stdev"], rhythmStats=["mean", "stdev"], tonalStats=["mean", "stdev"])("/dev/shm/'$core'.wav"); print("{:.2f}".format(fe["rhythm.bpm"]))')"; }; runfun
+
+
+# ok: 113   1off: 18   (131)   10off: 46   (177)   fail:  4   # 134 sec,  essentia/re2013
+# ok: 101   1off: 15   (116)   10off: 26   (142)   fail: 39   # bad-tags
+afun() { ffmpeg -hide_banner -v fatal -nostdin -ss $((dur/3)) -y -i /mnt/Users/ed/Music/"$fn" -ac 1 -ar 44100 /dev/shm/$core.wav || return 1; py="$(python3 -c 'from essentia.standard import *; a=MonoLoader(filename="/dev/shm/'$core'.wav")(); bpm,beats,confidence,_,intervals=RhythmExtractor2013(method="multifeature")(a); print("{:.2f}".format(bpm))')"; }; runfun
+
+
+
+########################################################################
+##
+##  key detectyion
+##
+########################################################################
+
+
+
+# console scriptlet reusing keytabs from browser.js
+var m=''; for (var a=0; a<24; a++) m += 's/\\|(' + maps["traktor_sharps"][a].trim() + "|" + maps["rekobo_classic"][a].trim() + "|" + maps["traktor_musical"][a].trim() + "|" + maps["traktor_open"][a].trim() + ')$/|' + maps["rekobo_alnum"][a].trim() + '/;'; console.log(m);
+
+
+# translate to camelot
+re='s/\|(B|B|B|6d)$/|1B/;s/\|(F#|F#|Gb|7d)$/|2B/;s/\|(C#|Db|Db|8d)$/|3B/;s/\|(G#|Ab|Ab|9d)$/|4B/;s/\|(D#|Eb|Eb|10d)$/|5B/;s/\|(A#|Bb|Bb|11d)$/|6B/;s/\|(F|F|F|12d)$/|7B/;s/\|(C|C|C|1d)$/|8B/;s/\|(G|G|G|2d)$/|9B/;s/\|(D|D|D|3d)$/|10B/;s/\|(A|A|A|4d)$/|11B/;s/\|(E|E|E|5d)$/|12B/;s/\|(G#m|Abm|Abm|6m)$/|1A/;s/\|(D#m|Ebm|Ebm|7m)$/|2A/;s/\|(A#m|Bbm|Bbm|8m)$/|3A/;s/\|(Fm|Fm|Fm|9m)$/|4A/;s/\|(Cm|Cm|Cm|10m)$/|5A/;s/\|(Gm|Gm|Gm|11m)$/|6A/;s/\|(Dm|Dm|Dm|12m)$/|7A/;s/\|(Am|Am|Am|1m)$/|8A/;s/\|(Em|Em|Em|2m)$/|9A/;s/\|(Bm|Bm|Bm|3m)$/|10A/;s/\|(F#m|F#m|Gbm|4m)$/|11A/;s/\|(C#m|Dbm|Dbm|5m)$/|12A/;'
+
+
+# runner/wrapper
+runfun() { cores=8; touch run; tbc() { bc | sed -r 's/(\.[0-9]{2}).*/\1/'; }; for ((core=0; core<$cores; core++)); do sqlite3 /mnt/Users/ed/Music/.hist/up2k.db 'select dur.w, dur.v, key.v from mt key join mt dur on key.w = dur.w where key.k = "key" and dur.k = ".dur" order by dur.w' | uniq -w16 | grep -vE '(Off-Key|None)$' | sed -r "s/ //g;$re" | uniq -w16 | while IFS=\| read w dur bpm; do sqlite3 /mnt/Users/ed/Music/.hist/up2k.db "select rd, fn from up where substr(w,1,16) = '$w'" | sed -r "s/^/$bpm /"; done| grep mir/cr | tr \| / | while read key fn; do [ -e run ] || break; n=$((n+1)); ncore=$((n%cores)); [ $ncore -eq $core ] || continue; t0=$(date +%s.%N); (afun || exit 1; t=$(date +%s.%N); td=$(echo "scale=3; $t - $t0" | tbc); [ "$key" = "$py" ] && c=2 || c=5; printf '%4s sec, %4s orig, \033[3%dm%4s py,\033[0m %s\n' $td "$key" $c "$py" "$fn") || break; done & done; time wait 2>/dev/null; }
+
+
+# ok: 26   1off: 10   2off: 1   fail: 3   #  15 sec, keyfinder
+afun() { ffmpeg -hide_banner -v fatal -nostdin -ss $((dur/3)) -y -i /mnt/Users/ed/Music/"$fn" -ac 1 -ar 44100 -t 60 /dev/shm/$core.wav || break; py="$(python3 -c 'import sys; import keyfinder; print(keyfinder.key(sys.argv[1]).camelot())' "/dev/shm/$core.wav")"; }; runfun
+
+
+# https://github.com/MTG/essentia/raw/master/src/examples/tutorial/example_key_by_steps_streaming.py
+# https://essentia.upf.edu/reference/std_Key.html  # edma edmm braw bgate
+sed -ri 's/^(key = Key\().*/\1profileType="bgate")/' example_key_by_steps_streaming.py
+afun() { ffmpeg -hide_banner -v fatal -nostdin -ss $((dur/3)) -y -i /mnt/Users/ed/Music/"$fn" -ac 1 -ar 44100 -t 60 /dev/shm/$core.wav || break; py="$(python3 example_key_by_steps_streaming.py /dev/shm/$core.{wav,yml} 2>/dev/null | sed -r "s/ major//;s/ minor/m/;s/^/|/;$re;s/.//")"; }; runfun
+
+
+
+########################################################################
+##
+##  misc
+##
+########################################################################
+
+
+
+python3 -m pip install --user vamp
+
+import librosa
+d, r = librosa.load('/dev/shm/0.wav')
+d.dtype
+# dtype('float32')
+d.shape
+# (1323000,)
+d
+# array([-1.9614939e-08,  1.8037968e-08, -1.4106059e-08, ...,
+#         1.2024145e-01,  2.7462116e-01,  1.6202132e-01], dtype=float32)
+
+
+
+import vamp
+c = vamp.collect(d, r, "vamp-example-plugins:fixedtempo")
+c
+# {'list': [{'timestamp':  0.005804988, 'duration':  9.999092971, 'label': '110.0 bpm', 'values': array([109.98116], dtype=float32)}]}
+
+
+
+ffmpeg -ss 48 -i /mnt/Users/ed/Music/mir/cr-a/'I Beg You(ths Bootleg).wav' -ac 1 -ar 22050 -f f32le -t 60 /dev/shm/f32.pcm
+
+import numpy as np
+f = open('/dev/shm/f32.pcm', 'rb')
+d = np.fromfile(f, dtype=np.float32)
+d
+array([-0.17803933, -0.27206388, -0.41586545, ..., -0.04940119,
+       -0.0267825 , -0.03564296], dtype=float32)
+
+d = np.reshape(d, [1, -1])
+d
+array([[-0.17803933, -0.27206388, -0.41586545, ..., -0.04940119,
+        -0.0267825 , -0.03564296]], dtype=float32)
+
+
+
+import vampyhost
+print("\n".join(vampyhost.list_plugins()))
+
+mvamp:marsyas_bextract_centroid
+mvamp:marsyas_bextract_lpcc
+mvamp:marsyas_bextract_lsp
+mvamp:marsyas_bextract_mfcc
+mvamp:marsyas_bextract_rolloff
+mvamp:marsyas_bextract_scf
+mvamp:marsyas_bextract_sfm
+mvamp:marsyas_bextract_zero_crossings
+mvamp:marsyas_ibt
+mvamp:zerocrossing
+qm-vamp-plugins:qm-adaptivespectrogram
+qm-vamp-plugins:qm-barbeattracker
+qm-vamp-plugins:qm-chromagram
+qm-vamp-plugins:qm-constantq
+qm-vamp-plugins:qm-dwt
+qm-vamp-plugins:qm-keydetector
+qm-vamp-plugins:qm-mfcc
+qm-vamp-plugins:qm-onsetdetector
+qm-vamp-plugins:qm-segmenter
+qm-vamp-plugins:qm-similarity
+qm-vamp-plugins:qm-tempotracker
+qm-vamp-plugins:qm-tonalchange
+qm-vamp-plugins:qm-transcription
+vamp-aubio:aubiomelenergy
+vamp-aubio:aubiomfcc
+vamp-aubio:aubionotes
+vamp-aubio:aubioonset
+vamp-aubio:aubiopitch
+vamp-aubio:aubiosilence
+vamp-aubio:aubiospecdesc
+vamp-aubio:aubiotempo
+vamp-example-plugins:amplitudefollower
+vamp-example-plugins:fixedtempo
+vamp-example-plugins:percussiononsets
+vamp-example-plugins:powerspectrum
+vamp-example-plugins:spectralcentroid
+vamp-example-plugins:zerocrossing
+vamp-rubberband:rubberband
+
+
+
+plug = vampyhost.load_plugin("vamp-example-plugins:fixedtempo", 22050, 0)
+plug.info
+{'apiVersion': 2, 'pluginVersion': 1, 'identifier': 'fixedtempo', 'name': 'Simple Fixed Tempo Estimator', 'description': 'Study a short section of audio and estimate its tempo, assuming the tempo is constant', 'maker': 'Vamp SDK Example Plugins', 'copyright': 'Code copyright 2008 Queen Mary, University of London.  Freely redistributable (BSD license)'}
+plug = vampyhost.load_plugin("qm-vamp-plugins:qm-tempotracker", 22050, 0)
+from pprint import pprint; pprint(plug.parameters)
+
+
+
+for c in plug.parameters: print("{} \033[36m{}  [\033[33m{}\033[36m] = {}\033[0m".format(c["identifier"], c["name"], "\033[36m, \033[33m".join(c["valueNames"]), c["valueNames"][int(c["defaultValue"])])) if "valueNames" in c else print("{} \033[36m{}  [\033[33m{}..{}\033[36m] = {}\033[0m".format(c["identifier"], c["name"], c["minValue"], c["maxValue"], c["defaultValue"]))
+
+
+
+beatroot-vamp:beatroot
+cl=c["list"]; 60*((len(cl)-1)/(float(cl[-1]["timestamp"]-cl[1]["timestamp"])))
+
+
+
+ffmpeg -ss 48 -i /mnt/Users/ed/Music/mir/cr-a/'I Beg You(ths Bootleg).wav' -ac 1 -ar 22050 -f f32le -t 60 /dev/shm/f32.pcm
+# 128 bpm, key 5A Cm
+
+import vamp
+import numpy as np
+f = open('/dev/shm/f32.pcm', 'rb')
+d = np.fromfile(f, dtype=np.float32)
+c = vamp.collect(d, 22050, "vamp-example-plugins:fixedtempo", parameters={"maxdflen":40})
+c["list"][0]["label"]
+# 127.6 bpm
+
+c = vamp.collect(d, 22050, "qm-vamp-plugins:qm-tempotracker", parameters={"inputtempo":150})
+print("\n".join([v["label"] for v in c["list"] if v["label"]]))
+v = [float(x["label"].split(' ')[0]) for x in c["list"] if x["label"]]
+v = list(sorted(v))[len(v)//4:-len(v)//4]
+v = sum(v) / len(v)
+# 128.1 bpm
+

docs/notes.sh

@@ -67,6 +67,33 @@ wget -S --header='Accept-Encoding: gzip' -U 'MSIE 6.0; SV1' http://127.0.0.1:392
 shab64() { sp=$1; f="$2"; v=0; sz=$(stat -c%s "$f"); while true; do w=$((v+sp*1024*1024)); printf $(tail -c +$((v+1)) "$f" | head -c $((w-v)) | sha512sum | cut -c-64 | sed -r 's/ .*//;s/(..)/\\x\1/g') | base64 -w0 | cut -c-43 | tr '+/' '-_'; v=$w; [ $v -lt $sz ] || break; done; }
 
 
+##
+## poll url for performance issues
+
+command -v gdate && date() { gdate "$@"; }; while true; do t=$(date +%s.%N); (time wget http://127.0.0.1:3923/?ls -qO- | jq -C '.files[]|{sz:.sz,ta:.tags.artist,tb:.tags.".bpm"}|del(.[]|select(.==null))' | awk -F\" '/"/{t[$2]++} END {for (k in t){v=t[k];p=sprintf("%" (v+1) "s",v);gsub(/ /,"#",p);printf "\033[36m%s\033[33m%s   ",k,p}}') 2>&1 | awk -v ts=$t 'NR==1{t1=$0} NR==2{sub(/.*0m/,"");sub(/s$/,"");t2=$0;c=2; if(t2>0.3){c=3} if(t2>0.8){c=1} } END{sub(/[0-9]{6}$/,"",ts);printf "%s   \033[3%dm%s   %s\033[0m\n",ts,c,t2,t1}'; sleep 0.1 || break; done
+
+
+##
+## sqlite3 stuff
+
+# find dupe metadata keys
+sqlite3 up2k.db 'select mt1.w, mt1.k, mt1.v, mt2.v from mt mt1 inner join mt mt2 on mt1.w = mt2.w where mt1.k = mt2.k and mt1.rowid != mt2.rowid'
+
+# partial reindex by deleting all tags for a list of files
+time sqlite3 up2k.db 'select mt1.w from mt mt1 inner join mt mt2 on mt1.w = mt2.w where mt1.k = +mt2.k and mt1.rowid != mt2.rowid'  > warks
+cat warks | while IFS= read -r x; do sqlite3 up2k.db "delete from mt where w = '$x'"; done
+
+
+##
+## media
+
+# split track into test files
+e=6; s=10; d=~/dev/copyparty/srv/aus; n=1; p=0; e=$((e*60)); rm -rf $d; mkdir $d; while true; do ffmpeg -hide_banner -ss $p -i 'nervous_testpilot - office.mp3' -c copy -t $s $d/$(printf %04d $n).mp3; n=$((n+1)); p=$((p+s)); [ $p -gt $e ] && break; done
+
+-v srv/aus:aus:r:ce2dsa:ce2ts:cmtp=fgsfds=bin/mtag/sleep.py
+sqlite3 .hist/up2k.db 'select * from mt where k="fgsfds" or k="t:mtp"' | tee /dev/stderr | wc -l
+
+
 ##
 ## vscode
 
@@ -96,6 +123,9 @@ for d in /usr /var; do find $d -type f -size +30M 2>/dev/null; done | while IFS=
 brew install python@2
 pip install virtualenv
 
+# readme toc
+cat README.md | awk '!/^#/{next} {lv=length($1);sub(/[^ ]+ /,"");bab=$0;gsub(/ /,"-",bab)} {printf "%" ((lv-1)*4+1) "s [%s](#%s)\n", "*",$0,bab}'
+
 
 ##
 ## http 206

scripts/make-sfx.sh

@@ -28,6 +28,13 @@ gtar=$(command -v gtar || command -v gnutar) || true
 	unexpand() { gunexpand "$@"; }
 	command -v grealpath >/dev/null &&
 		realpath() { grealpath "$@"; }
+
+	[ -e /opt/local/bin/bzip2 ] &&
+		bzip2() { /opt/local/bin/bzip2 "$@"; }
+}
+pybin=$(command -v python3 || command -v python) || {
+	echo need python
+	exit 1
 }
 
 [ -e copyparty/__main__.py ] || cd ..
@@ -38,11 +45,15 @@ gtar=$(command -v gtar || command -v gnutar) || true
 	exit 1
 }
 
+do_sh=1
+do_py=1
 while [ ! -z "$1" ]; do
 	[ "$1" = clean  ] && clean=1  && shift && continue
 	[ "$1" = re     ] && repack=1 && shift && continue
 	[ "$1" = no-ogv ] && no_ogv=1 && shift && continue
 	[ "$1" = no-cm  ] && no_cm=1  && shift && continue
+	[ "$1" = no-sh  ] && do_sh=   && shift && continue
+	[ "$1" = no-py  ] && do_py=   && shift && continue
 	break
 done
 
@@ -169,10 +180,11 @@ done
 	sed -r '/edit2">edit \(fancy/d' <$f >t && tmv "$f"
 }
 
+[ $repack ] ||
 find | grep -E '\.py$' |
   grep -vE '__version__' |
   tr '\n' '\0' |
-  xargs -0 python ../scripts/uncomment.py
+  xargs -0 $pybin ../scripts/uncomment.py
 
 f=dep-j2/jinja2/constants.py
 awk '/^LOREM_IPSUM_WORDS/{o=1;print "LOREM_IPSUM_WORDS = u\"a\"";next} !o; /"""/{o=0}' <$f >t
@@ -194,25 +206,36 @@ tar -cf tar "${args[@]}" --numeric-owner copyparty dep-j2
 
 echo compressing tar
 # detect best level; bzip2 -7 is usually better than -9
-for n in {2..9}; do cp tar t.$n; bzip2 -$n t.$n & done; wait; mv -v $(ls -1S t.*.bz2 | tail -n 1) tar.bz2
-for n in {2..9}; do cp tar t.$n;  xz -ze$n t.$n & done; wait; mv -v $(ls -1S t.*.xz  | tail -n 1) tar.xz
-rm t.*
+[ $do_py ] && { for n in {2..9}; do cp tar t.$n; bzip2 -$n t.$n & done; wait; mv -v $(ls -1S t.*.bz2 | tail -n 1) tar.bz2; }
+[ $do_sh ] && { for n in {2..9}; do cp tar t.$n;  xz -ze$n t.$n & done; wait; mv -v $(ls -1S t.*.xz  | tail -n 1) tar.xz; }
+rm t.* || true
+exts=()
 
+
+[ $do_sh ] && {
+exts+=(sh)
 echo creating unix sfx
 (
 	sed "s/PACK_TS/$ts/; s/PACK_HTS/$hts/; s/CPP_VER/$ver/" <../scripts/sfx.sh |
 	grep -E '^sfx_eof$' -B 9001;
 	cat tar.xz
 ) >$sfx_out.sh
+}
 
+
+[ $do_py ] && {
+exts+=(py)
 echo creating generic sfx
-python ../scripts/sfx.py --sfx-make tar.bz2 $ver $ts
+$pybin ../scripts/sfx.py --sfx-make tar.bz2 $ver $ts
 mv sfx.out $sfx_out.py
 chmod 755 $sfx_out.*
+}
+
 
 printf "done:\n"
-printf "  %s\n" "$(realpath $sfx_out)."{sh,py}
-# rm -rf *
+for ext in ${exts[@]}; do
+	printf "  %s\n" "$(realpath $sfx_out)."$ext
+done
 
-# tar -tvf ../sfx/tar | sed -r 's/(.* ....-..-.. ..:.. )(.*)/\2 `` \1/' | sort | sed -r 's/(.*) `` (.*)/\2 \1/'| less
-# for n in {1..9}; do tar -tf tar | grep -vE '/$' | sed -r 's/(.*)\.(.*)/\2.\1/' | sort | sed -r 's/([^\.]+)\.(.*)/\2.\1/' | tar -cT- | bzip2 -c$n | wc -c; done 
+# apk add bash python3 tar xz bzip2
+# while true; do ./make-sfx.sh; for f in ..//dist/copyparty-sfx.{sh,py}; do mv $f $f.$(wc -c <$f | awk '{print$1}'); done; done